2
Atmospheric Chemistry
Air quality, the health of the ozone layer, and the earth’s global climate areclosely tied to the composition of the atmosphere and the chemical transfor-mation of natural and anthropogenic trace gases. Although the purpose ofthis book is not to cover the full breadth of atmospheric chemistry, this chap-ter provides a short introduction into this topic, with the intent to motivateand clarify the applications of the differential optical absorption spectroscopy(DOAS) technique. Key elements of the contemporary understanding of at-mospheric chemistry and ongoing research efforts are presented. This chapteris mostly restricted to chemical reactions occurring between gas molecules,although a few surface reactions are also presented. It should be noted thatmuch of our knowledge of atmospheric chemistry comes from studies employ-ing sophisticated instrumentation for the detection of atmospheric trace gases,including the DOAS technique.
Much of the motivation to study atmospheric chemistry and the composi-tion of the atmosphere is based on various man-made environmental problemsthat have emerged over the past two centuries. Human activities have upsetthe natural balance of the atmosphere by influencing the trace gas and aerosolcomposition on local, regional, and even global scales. The following list namesthe most significant atmospheric environmental problems:
• ‘London Type’ smog was first recognised as an environmental problemin the 19th century (Brimblecombe and Heymann, 1998). The emissionof soot particles and the formation of aerosols consisting of sulphuricacid, which were produced from photochemical oxidation of SO2 emit-ted from combustion sources, had a severe impact on human health.During the so-called ‘London Killer Smog’, approximately 4000 excessdeaths were counted in London during a 4-day period. It should be notedthat the term ‘Smog’ was coined to describe the interaction of Smokeand Fog (Smog) that formed the strong haze observed in London duringwinter time.
U. Platt and J. Stutz, Atmospheric Chemistry. In: U. Platt and J. Stutz, Differential Optical
Absorption Spectroscopy, Physics of Earth and Space Environments, pp. 5–75 (2008)
DOI 10.1007/978-3-540-75776-4 2 c© Springer-Verlag Berlin Heidelberg 2008
6 2 Atmospheric Chemistry
• Crop damage and human discomfort were the first indication of high levelsof oxidants during sunny and hot days in Los Angeles (Haagen-Smit, 1952;Haagen-Smit and Fox, 1954; Finlayson-Pitts and Pitts, 2000). The forma-tion of this ‘Los Angeles Type’ smog was found to be primarily due tophotochemical formation of large amounts of ozone, carbonyl compounds,and organic aerosol from car exhaust and industrial emissions of nitrogenoxides, carbon monoxide, and volatile organic compounds (VOCs)(Haagen-Smit, 1952; Finlayson-Pitts and Pitts 2000). Despite manydecades of research and mitigation activities, Los Angeles type smog re-mains one of the most common air pollution problems in urban areastoday.
• Forest decline and lake acidification were signs of another man-made en-vironmental problem. Increased emissions of sulphur dioxide and nitrogenoxides, followed by their gas and aqueous phase oxidation to sulphuric andnitric acid, lead to ‘Acid rain’ that lead to the deposition of these acidsto various ecosystems (Charlson and Rhode, 1982). While acid rain hassuccessfully been reduced in Europe and the United States, it remains aproblem in many developing countries.
• The health of the atmospheric ozone layer has always been a concernof atmospheric scientists. In 1971, Johnston (1971) predicted the loss ofstratospheric ozone due to emission of oxides of nitrogen by a plannedfleet of supersonic passenger aircraft. Molina and Rowland (1974) warnedof a gradual loss of stratospheric ozone due to catalysed ozone destructionprocesses by halogen species transported to the stratosphere in the form ofextremely stable halogen-containing compounds, so called chlorofluorocar-bons (CFCs), used as coolants or spray can propellants. Their predictionsbecame reality in 1985, when Farman et al. (1985) discovered the strato-spheric ‘ozone hole’ over Antarctica. This recurring phenomenon reducesthe thickness of the stratospheric ozone layer over Antarctica to less thanone third of its normal level every Antarctic Spring (e.g. Farman et al.,1985; Solomon, 1999). While steps have been taken to stop the emissionsof CFCs, it is estimated that it will take another three decades before theozone hole will close again.
• One of the most serious environmental problems is the impact of humanactivities on the climate of our planet. The potential of global warmingcaused by IR-active ‘greenhouse’ gases such as CO2, CH4, N2O, O3, CFCs(e.g. CFCl3 and CF2Cl2), and the direct and indirect effects of chemicallygenerated aerosol, have been known for some time (e.g. Arrhenius, 1896;Rhode et al., 1997; IPCC, 2002). Human activities, and thus the emissionof the various greenhouse gases, over the past century have reached levelswhere an impact on global surface temperatures, the global water cycle,ocean levels, etc. has become very likely. Today, many signs point at thebeginning of a global climate change which will have a severe impact on ourearth. The global increase of tropospheric ozone, which apparently startedin the middle of the 19th century, also contributes to this phenomenon. Itis thought to be caused by catalytic ozone production due to wide-spread
2.1 Atmospheric Structure and Composition 7
emission of oxides of nitrogen and hydrocarbons (HC) (Volz and Kley,1988; Staehelin et al. 1994).
Many of these phenomena are closely related to the atmospheric chemistrypresented in this chapter.
Environmental conditions, such as temperature, pressure, and the solarspectrum, change with altitude and several altitude regimes of atmosphericchemistry can be distinguished. These properties of the atmosphere relevantfor atmospheric chemistry will be presented in Sect. 2.1. The composition ofthe atmosphere is influenced by emissions of natural and man-made species.A short overview over these emission and their sources is given in Sect. 2.2.Tropospheric chemistry is dominated by ozone and other oxidants, i.e. hy-droxyl/peroxy radicals, nitrogen oxides, and volatile organic carbons. Whiletheir chemistry is closely linked, we will discuss the chemistry of each class ofspecies in Sects. 2.4–2.7. Sulphur chemistry (Sect. 2.8) also plays an importantrole as it aids in the formation of atmospheric acid and is also crucial for theformation of particles in the atmosphere. The chemistry of reactive halogenspecies (RHS) in the troposphere, which is suspected to influence ozone levelson a global scale, is presented in Sect. 2.9. The various radical species in thetroposphere also play a crucial role for the ‘self-cleaning’ power of the atmo-sphere (Sect. 2.11). The atmospheric region between 15 and 50 km height, thestratosphere, hosts the ozone layer, which protects the earth surface from so-lar UV radiation. Consequently, stratospheric chemistry (Sect. 2.10) revolvesaround the formation and destruction pathways of ozone, as well as the chem-ical mechanisms leading to the destruction of ozone over Antarctica in spring,the so-called ozone hole.
2.1 Atmospheric Structure and Composition
For practical purposes, air can be viewed as an ideal gas. The relationshipbetween pressure p, absolute temperature T , and volume V for a given numberof moles n (1 mole is equal to NA = 6.023× 1023 molecules of air) is given bythe ideal gas law:
pV = nRT (2.1)
where R = 8.315 J mole−1K−1 denotes the universal gas constant.At standard sea-level pressure, p0 = 1.013 × 105 Nm−2 and temperature
T0 = 273.15K (or 0.0◦ Celsius), each cubic centimetre (cm3) of air contains2.69 × 1019 molecules. At higher altitudes, z, the atmospheric pressure, p(z)drops exponentially from its value at sea level:
p (z) = p0e−Mgz
RT = p0e− z
zS (2.2)
where M = 0.02897 kg mole−1 is the mean molar mass of air, g = 9.81 is theacceleration of gravity, and zS = RT
Mg ≈ 7 ± 1 km denotes the ‘scale height’ ofthe atmosphere.
8 2 Atmospheric Chemistry
While there is a monotonic drop in pressure with altitude, the verticaltemperature profile shows distinct variations, with several maxima and min-ima. The variation of temperature, type of chemical processes, and mixingmechanisms at different altitudes leads to a division of the atmosphere into anumber of distinct compartments (see Fig. 2.1):
The troposphere is the lowest layer of the atmosphere, and is relativelywell mixed. It extends from the surface to about 8 km altitude at the poles,and from the surface to about 18 km at the equator. The troposphere can befurther subdivided into:
• The atmospheric boundary layer (BL), which roughly covers the lowest1–2 km of the atmosphere. Air movement and mixing in the BL are influ-enced by friction on the earth’s surface.
• The free troposphere, which extends from the top of the BL to thetropopause.
The stratosphere is a poorly mixed layer extending above the troposphereto about 50 km altitude. Common subdivisions of the stratosphere include:
Fig. 2.1. The layers of the atmosphere, divided according to temperature, mixingmechanism, and degree of ionisation (from Brasseur and Solomon, 1986)
2.1 Atmospheric Structure and Composition 9
• The tropical tropopause layer• The lower and middle stratosphere• The upper stratosphere
The upper atmosphere extending above the stratopause is subdivided intothe mesosphere, thermosphere, and exosphere.
Despite the vertical temperature and pressure variation up to about 100 kmaltitude, the composition of the atmosphere is fairly constant with respectto its main components. The three most abundant species, which constitutemore than 99.9% of dry air (see Table 2.1), are molecular nitrogen and oxy-gen, as well as the noble gas argon. In addition, the atmosphere contains watervapour between 10−4% and a few % (by volume). Other species with compa-rable abundances are carbon dioxide, and all remaining nobel gases. With theexception of water vapour and carbon dioxide, the species listed in Table 2.1have atmospheric residence times in excess of 1000 years, and are thereforereferred to as ‘permanent’ constituents. In addition to the above species, theatmosphere contains a multitude of other species at much lower abundance.They are called ‘trace gases’ and include methane (CH4, ≈ 0.00017%), molec-ular hydrogen (H2, ≈ 0.00005%), and nitrous oxide (N2O, ≈ 0.00003%) amongothers.
2.1.1 Trace Species in the Atmosphere
In addition to the main gaseous constituents (Table 2.1), the atmospherecontains a large number of trace gases. Only a few trace species (e.g. CH4)exceed a mixing ratio of one molecule in a million air molecules. However, avery large number of species (e.g. ozone, carbonyl sulphide, difluoro–dichloro–methane, methyl chloride, and methyl bromide) are present at mixing ratiosaround or below one molecule in a billion air molecules. Atmospheric chemistry
Table 2.1. The main constituents of the (unpolluted), dry atmosphere
Gas Chemicalformula
Mixing ratioby volume
Mixing ratiovol. %
Nitrogen N2 0.7808 78.08Oxygen O2 0.2095 20.95Argon Ar 9.3 · 10−3 0.93Carbon dioxide CO2 0.37 · 10−3 0.037Neon Ne 18 · 10−6 0.0018Helium He 5.2 · 10−6 0.00052Methane CH4 1.7 · 10−6 0.00017Krypton Kr 1.1 · 10−6 0.00011Xenon Xe 0.9 · 10−6 0.00009Hydrogen H2 0.5 · 10−6 0.00005Dinitrogen oxide N2O 0.3 · 10−6 0.00003
10 2 Atmospheric Chemistry
is predominately concerned with the fate of these trace species, which, despitetheir low concentration, have a very noticeable impact on our atmosphere.
Another very important class of components in the atmosphere, which isnot gaseous, is small particles of liquid or solid matter dispersed and suspendedin air, the aerosol. The literal translation of the word aerosol is ‘solution in air’.Since a solution includes both a solvent and a solute, one might insist that theterm aerosol should only be used to denote the entire system, including boththe suspended particles and the carrier gas. Usually, however, the term aerosolis used in the literature synonymously with ‘suspended particles’, excludingthe carrier gas.
An important property of a solution is that it is a stable system. Applyinga similar criterion to aerosol particles means that only those particles whichremain suspended in air for a sufficiently long time to be transported by thewind over reasonably long distances can be part of an aerosol. The atmosphericresidence time of large aerosol particles is limited by their settling velocity,which is the cause of sedimentation. The settling velocity varies approximatelywith the square of the particle radius, and therefore the upper limit for thesize of aerosol particles is relatively sharp. Particles with radii much largerthan about 10 μm have too large settling velocities to behave like true aerosolparticles, and are thus classified as ‘coarse dust particles’. (A spherical particleof 10 μm diameter has a settling velocity of about 2.42 cm s−1, or 87mh−1,and thus will be rapidly removed from the atmosphere by sedimentation).
Defining the lower limit for the existence range of aerosol particles is moredifficult. Clearly, aerosol particles must exceed the size of individual molecules,which have radii in the range of fractions of a nanometre (e.g. the gas kineticradius of N2 or O2 is ≈ 0.2 nm). Nucleation theory teaches us that clusterscomposed of several condensable molecules (e.g. a mixture of sulphuric acidand water molecules) can only be stable when they exceed a certain criticalsize. Before reaching that critical size, the clusters tend to evaporate again. Itappears attractive to use that critical size as a lower limit for the definitionrange of aerosol particles in the atmosphere. However, the critical size is notuniquely defined, but depends on atmospheric conditions, in particular ontemperature, relative humidity, and the oxidation rate of SO2, which yieldsthe condensable sulphuric acid molecules. Very often the critical size of theseso-called ‘secondary aerosols’ is in the range of 1 nm.
2.1.2 Quantification of Gas Abundances
The amount of trace gases present in the atmosphere can be quantified byusing two different descriptions. First, we define the concentration of a tracegas as the amount of trace gas per volume of air (at a given temperature):
c =Amount of trace gas
Volume of air(2.3)
where ‘amount’ refers to either mass (cm), number of molecules (cn), or num-ber of moles (cM). Examples for units of concentration are micrograms per
2.1 Atmospheric Structure and Composition 11
m3 or molecules per cm3. The latter is also known as ‘number density’ of agas. The partial pressure of a species is also a measure of its concentration.
Second, we can define the mixing ratio of a trace gas as the ratio of theamount of a trace gas to the amount of air, including the trace gas:
x =Amount of trace gas
Amount of air + Trace gas≈ Amount of trace gas
Amount of air(2.4)
At typical atmospheric trace gas mixing ratios of <10−6, the distinction be-tween the ‘amount of air’ and ‘amount of air + trace gas’ is so small that itcan be neglected for practical purposes. It is necessary to distinguish whetherthe ‘amount’ is in volume, number of moles, number of molecules, or mass.
An example of a mixing ratio is parts per million (ppm) by volume:
xV =Unit volume of trace gas
106 unit volumes of (air + trace gas)ppm
For smaller trace gas mixing ratios, xV is given as parts per billion (ppb) andparts per trillion (ppt), which are analogously defined as:
xV =Unit volume of trace gas
109 unit volumes of (air + trace gas)ppb
xV =Unit volume of trace gas
1012 unit volumes of (air + trace gas)ppt
While mixing ratios could also be given by mass, this is rarely done in at-mospheric chemistry. Nevertheless, in the literature the terms ppmv, ppbv,and pptv are sometimes used to stress the fact that volume mixing ratios areunderstood. Similar to volume mixing ratios are molar mixing ratios:
xM =Moles of trace gas
Mole of (air + trace gas)(2.5)
It should be noted that xM is the new SI unit for mixing ratios. Since air underambient conditions can be regarded in good approximation as an ideal gas, forpractical purposes a mixing ratio xM specified in moles per mole equals themixing ratio xV. Thus, the terms micromole per mole, nanomole per mole,and picomole per mole are essentially equivalent to ppm, ppb, and ppt byvolume, respectively.
A common problem in atmospheric chemistry is the conversion of unitsof the amounts of trace gases. For example, for trace gas i we obtain thefollowing conversion between the number density cn (in molecules per cm3)and mass concentration cm (in grams per cm3):
(cm)i =cn · Mi
NA(2.6)
where Mi denotes the molecular mass of the species i in g per mole and NA
is Avogadro’s number with NA = 6.0221420 × 1023 molecules mole−1. Thus,the conversion between number density and mass concentration is differentfor species with different molecular (atomic) mass.
12 2 Atmospheric Chemistry
The conversion of number density cn (in molecules per cm3) into the cor-responding (volume) mixing ratio is given by:
xV = cnV0
NAor cn = xV · NA
V0(2.7)
where V0 denotes the molar volume in cm3 for the pressure p and temperatureT at which the number density cn was measured. For standard conditions(p0 = 101325Pa = 1 Atmosphere, T = 273.15K) the molar volume is V0 =22414.00 cm3 mole−1. For arbitrary temperature and pressure conditions, wecan use:
xV = cn1
NA
T
p
p0
T0V0 = cn
1NA
· RT
p= cm
1Mi
· RT
p(2.8)
or
cm = xV · Mi ·p
RT(2.9)
Table 2.2 gives a few examples for the conversion of the different units inwhich the abundance of trace gases is customarily expressed.
It should be noted that mixing ratios, xV and xM, are independent oftemperature and pressure. Consequently, they are conserved during verticaltransport in the atmosphere. In contrast, the concentration c depends on bothparameters and change when air is transported. However, trace gas concentra-tions are relevant for the calculation of chemical reaction rates and radiativeproperties (such as UV absorption) of the atmosphere. Spectroscopic mea-surement techniques (including DOAS) also give results in number density orconcentration, not mixing ratios.
Table 2.2. The different units for the abundance of atmospheric trace gases(T = 293.15K, p = 101325Pa)
Trace gas Molecularmass g/mole
Mixing ratioxV ppb
Numberdensitycn molec. cm−3
Concentrationcm μg m3
O3 48.00 1.000 2.503 · 1010 1.9950.501 1.254 · 1010 1.000
SO2 64.06 1.000 2.503 · 1010 2.6620.376 0.941 · 1010 1.000
NO 30.01 1.000 2.503 · 1010 1.2510.799 2.000 · 1010 1.000
NO2 46.01 1.000 2.503 · 1010 1.9120.532 1.33 · 1010 1.000
CH4 16.04 1.000 2.503 · 1010 0.6671.500 3.755 · 1010 1.000
CH2O 30.03 1.000 2.503 · 1010 1.2480.801 2.005 · 1010 1.000
CO 28.01 1.000 2.503 · 1010 1.1640.859 2.150 · 1010 1.000
2.1 Atmospheric Structure and Composition 13
Thus both types of units are useful and must be converted into each otheras needed. Atmospheric trace gas profiles might look quite different whenviewed in terms of mixing ratios or concentrations, as illustrated by the ex-ample of the atmospheric ozone profile in Fig. 2.2. The concentration (inmolecules per cm3) has a noticeable, essentially constant level in the tropo-sphere, and a maximum at about 22 km in the lower stratosphere. In contrast,tropospheric mixing ratios (in ppm or μmole per mole) are very low and peakat about 36 km.
Another category for specifying trace gas abundances is the column densityS, which is defined as the integral of the number density along a certain pathin the atmosphere:
S = Sn =∫
Path
cn (s) ds (2.10)
The unit of Sn, and column densities in general, is molecules per cm2. Inanalogy to the mass mixing ratio cm, the mass column density Sm can bedefined as:
Sm =∫
Path
cm (s) ds (2.11)
Typical units for mass column densities are μg/cm2. The conversion betweenSn and Sm is analogous to the conversion between cn and cm.
0
0 2 4 6 8 10
ppm
0
10
20
30
40
50
60
1 × 1012
Hei
ght (
km)
2 × 1012 3 × 1012
Concentration molecules (cm–3)
4 × 1012 5 × 1012
Fig. 2.2. Vertical profile of ozone in troposphere and stratosphere (US StandardAtmosphere), both in mixing ratio (in ppm or μmole per mole), peaking at about36 km and concentration (molecules per cm3), peaking at about 22 km
14 2 Atmospheric Chemistry
A frequently used path is the total (vertical) atmospheric column den-sity V , where the concentration is integrated vertically from the surface toinfinity:
V =
∞∫
0
cn (z) dz (2.12)
2.1.3 Lifetime of Trace Gases in the Atmosphere
The atmosphere is not a static system with respect to its components. Rather,gases are released to the atmosphere or formed by chemical transformations,and removed again by chemical degradation or deposition on the ground.A useful quantity in this context is the steady-state (average) lifetime, orresidence time, of a species in the atmosphere.
If the concentration of a species has reached a steady-state value, i.e. itsrate of production P equals its rate of destruction D, its lifetime can be definedas the ratio:
τ =c
D=
c
P(2.13)
Assuming the production P to be constant and the destruction D to be pro-portional to c, i.e. D = c/τ with the constant of proportionality 1/τ , weobtain:
c (t) = P · τ ·(1 − e−
tτ
)−−−−→τ→∞
P · τ (2.14)
The constant τ becomes clear by calculating c(τ) = 1 − 1/e ≈ 63.2% ofthe final value c(∞) = c(t → ∞) = P · τ . In other words, the lifetime τis the time constant with which the trace gas concentration approaches itssteady state.
Figure 2.3 gives approximate lifetimes of atmospheric trace constituents.The abundance of atmospheric constituents frequently is in a stationary statewhich depends on the lifetime of the particular gas. The variation of the con-centration of a gas in space and time also depends on its lifetime. For instance,the long-lived components such as N2, O2, or noble gases, with lifetimes ofthousands of years, show very little variation, whereas the concentrations ofshort-lived species may vary considerably with time and location. A quanti-tative relationship of residence time and variation of the mixing ratio of a gaswas first established by Junge (1963).
2.2 Direct Emission of Trace Gasesto the Atmosphere
As described above (Sect. 2.1.3), most gaseous (and particulate) compo-nents of our atmosphere are in a balance between continuous addition to
2.2 Direct Emission of Trace Gases to the Atmosphere 15
Micro Local Regional Global
• CFC’s• N2O
• CH4
• CH3CCl3• CH3Br
• CO
• Trop O3
• SO2
• NOx• DMS• C3H6
• C5H8
Mod
erat
ely
Long
-Liv
edS
hort
-Liv
ed• CH3O2• HO2
• NO3
• OH
1m 10m 100m 1km 10km 100km1000km 10,000km
Transport Scale
Long
-Liv
ed
1s
100
s
1hr
1 d
ay
1yr
1
0yr
s 1
00 y
rs
Spatial Scale
Life
time
/ Res
iden
ce T
ime
Fig. 2.3. Average lifetimes (residence times) of gases in the atmosphere rangefrom seconds (and below) to millennia. Accordingly, transport can occur onscales reaching from a few meters to the global scale (adapted from Seinfeld andPandis, 1998)
the atmosphere and removal from it. To understand the levels of trace gasesin the atmosphere and their chemistry, it is important to study their sourcesor, in the case of gases which are formed chemically, the sources of their pre-cursors. In particular with respect to air pollution, this is an important issuesince human activities have added new trace gas sources over the past cen-turies. These sources have to be compared with natural emission sources inorder to assess their impact on the composition of the atmosphere. While thissection only gives a very brief overview of trace gas sources, which is lim-ited to the gases that are of interest in this book, it should be noted thatthe accurate determination of emissions is a crucial aspect of atmosphericchemistry.
Emission sources are typically subdivided into natural and anthropogenicprocesses. The following list gives examples of the types of sources found inthe atmosphere:
16 2 Atmospheric Chemistry
Natural emission of gases includes the following processes:
• Emission of gases from soil (e.g. methane, oxides of nitrogen NO and N2O)• Emission of gases by vegetation (primarily volatile organic compounds
(VOC))• Emission by biomass burning (VOCs, CO, NOx, etc.)• Marine emissions [Dimethyl sulphide (DMS) and sea salt aerosol]• Volcanic emissions (CO2, SO2, HCl, BrO, etc.)• Formation in thunderstorms (NO)
Anthropogenic emissions are due to a series of human activities, primarilytraffic, heating, and industrial processes, and also agricultural activities. Thelatter include emission from bare, or fertilised soils, flooded areas (e.g. for ricegrowing), or biomass burning. Anthropogenic emissions include:
• Emission during combustion processes (e.g. cars, power stations, and in-dustry). The emitted gases include oxides of nitrogen (NO and NO2),carbon monoxide, and VOCs
• Emission by industrial processes
Some anthropogenic emissions occur through the same processes as naturalemission:
• Impact on gas emissions from soil due to agriculture: oxides of nitrogen(NO and N2O)
• Emission by biomass burning (VOCs, CO, and NOX)• Emission of gases, by vegetation (primarily VOC)
We will briefly discuss the source and sink strengths, as well as average mix-ing ratios and lifetimes (residence times), of a series of important atmospherictrace gases, which have their dominant sources at the surface. The discussionis subdivided into nitrogen, sulphur, and carbon-containing species.
2.2.1 Nitrogen Species
Fixed nitrogen species play an important role in the chemistry and the climateof the atmosphere. We will focus here on the three most nitrogen importantspecies that are emitted into the atmosphere – nitrous oxide (N2O), nitrogenoxides (NO and NO2), and ammonia (NH3).
N2O
Nitrous oxide is an important greenhouse gas and also the dominant source ofreactive nitrogen in the stratosphere. N2O is naturally emitted by bacteria inthe soil and the ocean during the nitrogen fixation process (Table 2.3). Humanactivities have lead to an increase of these sources, predominantly due to theintensivation and expansion of agriculture. As a consequence, N2O mixingratios have increased from ∼275 ppb before the year 1800 to over 310 ppbtoday.
2.2 Direct Emission of Trace Gases to the Atmosphere 17
Table
2.3
.Tro
posp
her
icso
urc
es,si
nks,
and
mix
ing
rati
os
ofnit
rogen
spec
ies
[Sourc
est
rength
inm
illion
tons
offixed
nit
rogen
(Tg
N)
per
yea
r(f
rom
Lee
etal.,1997,E
hhalt
1999)]
Spec
ies
Mix
ing
rati
o,
ppb
Majo
rso
urc
esA
ver
age
sourc
est
rength
Tg/yea
rM
ajo
rSin
ks
Sin
kst
rength
Tg/yea
rA
tm.life
tim
e
Nit
rous
oxid
e,N
2O
310
Soil
emis
sion
Oce
an
emis
sion
Anth
ropogen
ic
10
6 9
Loss
tost
rato
spher
e25
110
a
Nit
rogen
oxid
es,N
O,
NO
2
0.0
3–5
Foss
ilfu
elco
mbust
ion
Bio
mass
burn
ing
Soil
emis
sion/m
icro
bia
lpro
duct
ion
Thunder
storm
sA
ircr
aft
emis
sions
21
8 10
5 0.8
Oxid
ati
on
by
OH
and
O3
toH
NO
3
44
2d
Am
monia
,N
H3
0.1
mari
ne
5co
nti
nen
tD
om
esti
canim
als
Em
issi
on
from
veg
etati
on
Oce
an
emis
sion
Fer
tilize
ruse
26
6 9 8
Dry
dep
osi
tion
Conver
sion
toN
H4+
aer
oso
l
17
36
5d
h=
hour,
d=
day
,a
=yea
r.
18 2 Atmospheric Chemistry
NOX (NO + NO2)
The significance of nitrogen oxides will be discussed in detail in Sect. 2.5. Ox-ides of nitrogen, NO and NO2, are produced in a large number of natural andanthropogenic processes. Particularly important are those where air is heatedto high temperatures, such as in combustion processes (in internal combustionengines), forest fires, or lightning strikes. Natural sources of nitrogen oxidesproduce globally about 13–31 million tons of nitrogen fixed as NOX per year(Table 2.3). The largest natural contributions come from brush and forestfires, lightning storms, and emission from the soil. Smaller contributions comefrom diffusion from the stratosphere and oxidation of ammonia (Ehhalt andDrummond, 1982; Lee et al., 1997).
Anthropogenic NOX emissions can be grouped into three categories thatessentially all originate from combustion processes. Thus, the largest sourceis the stationary combustion of fossil fuel (power stations, industry, and homeheating). A further important – and growing – source is emission by automo-biles. This emission additionally occurs in densely populated areas at very lowemission height. Finally, part of the NOX emission from forest fires and fromsoil is due to anthropogenic influence (intentional forest fires and artificialfertilisation).
In Table 2.3, the contributions of the various NOX sources are summarised.On a global scale, the contributions of natural and anthropogenic NOX sourcesare about equal in magnitude. The spatial distribution of strong NOX sources,as derived from satellite-based DOAS measurements, is shown in Figs. 11.45and 11.46 (Leue et al., 2001; Beirle et al., 2004a,c).
It should, however, be considered that the natural NOX sources are muchmore equally distributed over the surface of the earth than the anthropogenicsources, which are concentrated on a very small fraction of the earth’s sur-face (see also Fig. 11.45). For instance, in Germany [NOX emission 2000: 0.9million tons nitrogen, as N (Schmolling, 1983; Fricke and Beilke, 1993)], thenatural contribution to the total NOX emission is small.
NH3
Ammonia plays a crucial role in the atmosphere as it is the only significantatmospheric base neutralising acids such as sulphuric and nitric acids. Theammonium ion, which is formed upon the uptake of ammonia on particles,is an important part of the aerosol. Ammonia is emitted predominately bylivestock wastes and fertilised soils. Emissions from the ocean and vegetationalso play a role (Table 2.3). Modern cars with catalytic converters also emitammonia.
2.2.2 Sulphur Species
The increase of sulphur emissions since the onset of industrialisation has ledto numerous environmental problems, such as London-type smog and acid
2.2 Direct Emission of Trace Gases to the Atmosphere 19
rain. Naturally sulphur is emitted through biological processes in the soil andthe ocean in its reduced form as carbonyl sulphide (COS), hydrogen sulphide(H2S), and DMS (CH3SCH3). Volcano eruptions can also contribute to thenatural sulphur emissions, in the form of sulphur dioxide, SO2 (Table 2.4). Ofthe natural emissions, those of DMS dominate and are particularly importantfor the global sulfur budget (Sect. 2.7).
Anthropogenic emissions of SO2 originate primarily from fossil fuel com-bustion. Over the past 200 years, sulphur emissions have sharply increased.As a response to the growing acid rain problem, the emission of sulphur hasbeen greatly reduced in industrialised countries since 1975 (see also Fig. 2.4).This is due to stack gas desulphurisation measures, and also to the changeof the economies in the eastern European countries. On a global scale, how-ever, there is an expected increase of sulphur emission, due to enhanced coalcombustion in Asian countries (Fig. 2.4).
The sulphur species that are most important in the atmosphere, theirtypical concentrations and atmospheric residence times, and their degradationmechanisms are summarised in Table 2.4.
2.2.3 Carbon-Containing Species
A very large number of carbon-containing trace species are present in theatmosphere. The most important carbon-containing species is carbon dioxide(CO2). While CO2 plays a crucial role as a greenhouse gas, it undergoes littlechemistry in the atmosphere, and its sources will not be a topic of this section.The second most abundant carbon species is methane (CH4), which, as willbe discussed in Sect. 2.4, is chemically degraded in the atmosphere, directlyparticipating in the formation of tropospheric ozone. Methane is also a green-house gas. It is difficult to distinguish natural and anthropogenic sources ofmethane as the emissions predominately stem from animals, including farmlivestock, and wetlands, which includes rice paddies (Table 2.5). Fossil fuelconsumption also contributes to today’s methane emissions. The expansionof agriculture over the past two centuries has led to an increase of methanesources and a rise of methane mixing ratios from ∼750 ppb to over 1700 ppbtoday. Carbon monoxide (CO) is a product of incomplete combustion of nat-ural material, for example forest fires, and fossil fuels. In addition, CO is aproduct of the chemical degradation of methane and volatile organic carbon(VOC) species in the atmosphere. Because CO is toxic and participates in theformation of photochemical smog, various measures have been taken to reduceits emissions from combustion sources. The most prominent example is thecatalytic converter of automobiles, which was primarily designed to reduceCO emissions.
Hydrocrabons – often the term non-methane hydrocarbons (NMHC) isused to exclude CH4 from this group – constitute a large class of carbon-containing species in the atmosphere. The number of different species is verylarge and we will consider only the most important members of this group
20 2 Atmospheric Chemistry
Table
2.4
.Tro
posp
her
icso
urc
es,si
nks,
and
mix
ing
rati
os
of
sulp
hur
spec
ies
[Sourc
est
rength
inm
illion
tons
of
sulp
hur
(Tg
S)
per
yea
r(f
rom
Ehhalt
1999)]
Spec
ies
Mix
ing
rati
o,
ppb
Majo
rso
urc
esSourc
est
rength
Tg/yea
rM
ajo
rsi
nks
Sin
kst
rength
Tg/yea
rA
tm.life
tim
e
Carb
onyl
sulp
hid
e,C
OS
0.5
Soil
emis
sion
0.3
Upta
ke
by
veg
etati
on
0.4
7a
Oce
an
emis
sion
0.3
Flu
xto
stra
tosp
her
e0.1
5C
S2
+D
MS
ox.
0.5
Hydro
gen
sulp
hid
e,H
2S
0.0
05–0.0
9Soil
emis
sion
Veg
etati
on
Volc
anoes
0.5
1.2
1
Rea
ctio
nw
ith
OH
2.7
3d
Dim
ethyl
sulp
hid
e,D
MS,
CH
3SC
H3
0.0
05–0.1
Oce
an
emis
sion
Soil
emis
sion
68
2R
eact
ion
wit
hO
H,
NO
3,B
rO70
2d
Sulp
hur
dio
xid
e,SO
2
0.0
2–0.0
9(m
ari
ne)
,0.1
–5
cont.
Foss
ilfu
elburn
ing
160
Dry
dep
osi
tion,
react
ion
wit
hO
H,
liquid
phase
100
4d
Volc
anoes
14
Oxid
ati
on
toSO
4=
140
Sulp
hid
eox
idati
on
70
+W
etdep
osi
tion
h=
hour,
d=
day
,a
=yea
r.
2.3 Ozone in the Troposphere 21
1980
(Gg)
B
5000
0
10000
15000
20000
25000
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
Africa
Ocean
S. AmericaMid East
N. AmericaW. Europe
Asia
E. Europe
Fig. 2.4. Regional trends of global sulphur emissions (in giga-grams of sulphur peryear) during the past two decades (from Stern, 2005, Copyright Elsevier, 2005)
here. On a global scale, the emission by vegetation of isoprene, C5H8, andto a lesser extent, terpenes, C10H16, dominate the emissions of VOC. Thesespecies are highly reactive in the atmosphere and participate in the formationof ozone and particles. Globally, natural emissions of other species are of lesserimportance. Anthropogenic (and natural) emissions of VOC are an essentialpart of Los Angeles-type smog. A complex mixture of different trace gasesis emitted by fossil fuel consumption, both through incomplete combustionand the loss of VOC vapours before combustion, industrial processes, refin-ing of oil, solvents, etc. The emitted mixture, which contains alkanes (ethane,propane, etc.), alkenes (ethane, propene, etc.), alkynes and aromatics (ben-zene, toluene, etc.), and a variety of oxidised VOC (formaldehyde, acetalde-hyde, etc.), depends on the emissions signatures of the different sources, as wellas the distribution of these sources. For details on urban VOC emissions, werefer the reader to atmospheric chemistry textbooks, such as Finalyson-Pittsand Pitts (2000).
2.3 Ozone in the Troposphere
Ozone is a key compound in the chemistry of the atmosphere. In the tropo-sphere it is a component of smog, poisonous to humans, animals and plants, aswell as a precursor to cleansing agents (such as the OH radical, see Sect. 2.4.1).Tropospheric ozone is also an important greenhouse gas.
22 2 Atmospheric Chemistry
Table
2.5
.Im
port
ant
org
anic
trace
gase
s,appro
xim
ate
mix
ing
rati
o,em
issi
on
toth
eatm
osp
her
e,re
mov
alfr
om
the
atm
osp
her
e,and
aver
age
life
tim
e(r
esid
ence
tim
e)τ
inth
eatm
osp
her
e[s
ee(2
.14)]
Spec
ies
Mix
ing
rati
o,ppb
Majo
rso
urc
esSourc
est
rength
Tg/yea
r
Majo
rSin
ks
Sin
kst
rength
Tg/yea
r
Atm
.life
tim
e
Met
hane,
CH
41700
Ric
efiel
ds
75
Rea
ctio
nw
ith
OH
490
8a.
Dom
esti
canim
als
100
Export
tost
rato
spher
e40
Sw
am
ps/
mars
hes
200
Soil
upta
ke
30
Bio
mass
burn
ing
40
Foss
ilfu
el100
Carb
on
monox
ide,
CO
200
(NH
)50
(SH
)A
nth
ropogen
icem
issi
on
440
Rea
ctio
nw
ith
OH
2400
0.2
a
Bio
mass
burn
ing
770
Flu
xto
stra
tosp
her
e100
CH
4ox
idati
on
860
Soil
upta
ke
400
Ox.ofnatu
ralV
OC
610
Isopre
ne,
C5H
80.6
–2.5
Em
issi
on
from
dec
iduous
tree
s570
Rea
ctio
nw
ith
OH
,ozo
noly
sis
570
0.2
d
Ter
pen
es,C
10H
16
0.0
3–2
Em
issi
on
from
conifer
ous
tree
s140
Rea
ctio
nw
ith
OH
,ozo
noly
sis
140
0.4
d
Tota
lV
OC
Foss
ilfu
el65
Rea
ctio
nw
ith
OH
850
Bio
mass
burn
ing
35
Foliar
emis
sion
744
Oce
an
emis
sion
6
The
table
isbase
don
Ehhalt
(1999),
but
has
bee
nex
tended
.h
=hour,
d=
day
,a
=yea
r.
2.3 Ozone in the Troposphere 23
Ozone is formed by two distinctly different mechanisms in the troposphereand stratosphere. In the stratosphere, O2 molecules are split by short-waveUV radiation into O-atoms, which combine with O2 to form O3. This pro-cess is the core of the ‘Chapman Cycle’ (Chapman, 1930; see Sect. 2.10.1).Until the late 1960s, it was believed that tropospheric ozone originated fromthe stratosphere. Today we know that large amounts of O3 are formed anddestroyed in the troposphere. Influx of O3 from the stratosphere is only a mi-nor contribution to the tropospheric ozone budget. Recent model calculations(IPCC, 2002) put the cross tropopause flux of O3 at 390–1440 Mt a−1 (veryrecent investigations indicate that values near the lower boundary of the rangeare more likely), while they derive ozone formation rates in the troposphereat 2830–4320 Mt a−1. The formation is largely balanced by photochemicaldestruction in the troposphere amounting to 2510–4070 Mt a−1. Another, rel-atively small contribution to the O3 loss is deposition to the ground, modelledat 530–900 Mt a−1.
2.3.1 Mechanism of Tropospheric Ozone Formation
In the early 1950s, it became clear that under certain conditions in the atmo-sphere near the ground high concentrations of ozone are formed (e.g. Haagen-Smit, 1952). In fact, it could be shown in ‘smog-chamber’ experiments thatlarge amounts of ozone are formed when mixtures of NOX (NO + NO2) andVOC are exposed to solar UV radiation. While the phenomenon of ozoneformation as a function of VOC and NOX in illuminated mixtures was empir-ically found in the 1960s, the exact mechanism only became clear much later,due to the work of Weinstock (1969), Crutzen (1970), and Levy et al. (1971).Ozone formation in the troposphere is initiated by the production of O(3P)from NO2 photolysis (indicated by the term hν, see Chap. 3).
Under clear sky conditions at noontime, the average lifetime of the NO2
molecule is only on the order of 2 min [jNO2 = j2 ≈ 8× 10−3s−1; e.g. Junker-mann et al. (1989)]:
NO2 + hν → NO + O(3P) (R2.1)This reaction is followed by the rapid recombination of O with O2:
O(3P) + O2 + M → O3 + M (R2.2)
Where M denotes any atmospheric molecule. At high pressure (and thusM and O2 concentrations) in the troposphere, other reactions of O(3P), inparticular its reaction with O3, are negligible. Therefore, for each photolysedNO2 molecule, an ozone molecule is formed. Reactions R2.1 and R2.2 areessentially the only source of ozone in the troposphere. However, ozone isoften rapidly oxidised by NO to back NO2:
O3 + NO → NO2 + O2 (R2.3)
Reactions R2.1–R2.3 lead to a ‘photo-stationary’ state between O3, NO, andNO2. The relation between the three species can be expressed by the LeightonRelationship (Leighton, 1961):
24 2 Atmospheric Chemistry
[NO][NO2]
=j2.1
k2.3 · [O3](2.15)
where j2.1 denotes the photolysis frequency of NO2 and k2.2 denotes the rateconstant for the reaction of ozone with NO. For typical ozone mixing ratiosof 30 ppb (1 ppb ≈ 10−9 mixing ratio), the [NO]/[NO2] ratio during day-time near the ground is on the order of unity. The reaction cycle formed byR2.1–R2.3 does not lead to a net formation of ozone. However, any reactionthat converts NO into NO2 without converting an O3 molecule interferes withthis cycle and leads to net ozone production. The key factor in troposphericO3 formation is, thus, the chemical conversion of NO to NO2.
In the troposphere, the conversion of NO to NO2 without O3 occursthrough a combination of the reaction cycles of hydroxyl HOX (OH + HO2),peroxy radicals, and NOX (Fig. 2.5). In these cycles, OH radicals are con-verted to HO2 or RO2 radicals, through their reaction with CO or HC. HO2
and RO2, on the other hand, react with NO to reform OH, thus closing theHOX/ROX cycle. This reaction also converts NO to NO2 (see also Sect. 2.5.1),which is then photolysed back to NO (R2.1). The oxygen atom formed inthe NO2 photolysis then reacts with O2 to form ozone (R2.2). The process,shown in Fig. 2.5, therefore acts like a machine that, in the presence of NOxand sunlight, converts the ‘fuel’ CO and HC into CO2, water, and ozone.Because, HOX and NOX are recycled, this catalytic ozone formation can
NO2
O3+hν+H2O
HNO3
Fig. 2.5. Ozone formation in the troposphere is catalysed by hydrogen radicals(OH + HO2=HOX), peroxy radicals, and NOX
2.3 Ozone in the Troposphere 25
be quite efficient (Crutzen, 1970). HOX radicals are always present in thesunlit atmosphere (see Sect. 2.4); they are, for example, formed through thephotolysis of ozone in the presence of water vapour. The cycles are only in-terrupted if either a NOX or a HOX is removed from the respective cycles,for example by the reaction of OH with NO2, or the self-reactions of HO2 andRO2.
In background air, fuel for ozone formation is always present in the form ofmethane (mixing ratio of ∼1.7 ppm) and CO, which is formed as a degradationproduct of CH4 (Sect. 2.4.1). However, in clean air, the NOX level might bevery low, and thus insufficient to act as catalyst. This effect is outlined inmore detail in Sect. 2.5.1.
2.3.2 Ozone Formation in Urban Centresand Downwind
In many urban centres, car exhaust dominates as a source of pollution. Emis-sions from home heating and industry can also contribute considerably. Thesesources emit large amounts of NOX (mostly in the form of NO) and VOC,the main ingredients of ozone formation. As a consequence, ozone formationrates can be orders of magnitude higher than in slightly polluted air outsideurban centres. On the other hand, in urban centres there are also effectiveozone sinks:
• Reaction of O3 with freshly emitted NO• Reaction of O3 with olefins.
Overall, the ozone lifetime can be quite short (on the order of a few hours)in urban areas. Therefore, there are usually large diurnal variations of the O3
level in these regions. In fact, in urban centres the ozone level near the groundfrequently reaches zero a few hours after sunset and remains at zero for therest of the night. After sunrise, enhanced vertical mixing replenishes some O3
from higher layers, thus initiating radical chemistry again (see Fig. 2.8). Inaddition, photolysis of formaldehyde and nitrous acid (see Sects. 2.4.1 and2.5.1) can considerably speed up radical processes, and thus growth of the O3
mixing ratio in the morning.In a series of experiments where the concentrations of NOX and VOC
were independently varied, the ozone formation after a fixed time was foundto vary with the initial levels of both groups of species in a characteristic way:At a given initial level of NOX , the O3 increases linearly with the initiallypresent level of VOC. In other experiments with higher initial VOC levels,the amount of O3 levels off. Increasing initial VOC even further does not leadto higher O3 production. This regime is called ‘NOX limited’. Conversely,at a given initial level of VOC, the O3 increases linearly with the initiallypresent level of NOX . At higher initial NOX levels, the amount of O3 levelsoff. Increasing initial NOX even further in this ‘VOC limited’ regime does
26 2 Atmospheric Chemistry
00
1
2
3
4
5
6
2 4 6 8 10
NOx emissions, 1011 molec. cm–2 s–1
Hyd
roca
rbon
em
issi
ons,
1011
ato
ms
C c
m–2
s–1
12 14 16 18
20406080
100
120
140
160
180
200160
140
12010080
60
Fig. 2.6. Ozone concentrations simulated by a regional photochemical model as afunction of NOX and hydrocarbon (VOC) emissions. The thick line separates theNOX limited regime (top left) from the VOC limited regime (bottom right) (fromSillman et al., 1990)
not lead to higher O3 production. In fact, due to removal of OH by reactionwith NO2 (Fig. 2.5), O3 formation will be reduced. Figure 2.6 illustrates thisrelationship. The lines of constant ozone formation as function of the VOCand NOX composition of the initial mixture (or VOC and NOX emission) arecalled ozone isopleths. Measurements in the open atmosphere also illustratethis dependence of the ozone production on the NOX level (an example isshown in Fig. 2.7).
When considering which ‘mixture’ of HC (VOC) and NOX is most effectivein ozone production (at a given level of solar UV), it turns out that NOX andVOC must be present in a certain ratio RX0 = [NOX ]/[VOC], indicated bythe ridge line in Fig. 2.6. Often this maximum is not reached in urban centres,where the primary emissions occur. Because the atmospheric lifetimes of VOCand NOX are different, RX0 for an air-mass changes with time. In addition,it takes several hours for high ozone levels to be formed. Consequently, thehighest O3 levels are often found downwind of urban centres. This is illustratedin Fig. 2.8 which shows the diurnal variation of ozone levels in different typesof air masses at a rural site (Forest, Weltzheimer Wald, Germany), and in acity (Heilbronn, southern Germany).
2.4 Radical Processes in the Atmosphere 27
Fig. 2.7. Ozone production rates (PO3) calculated from simultaneous observationsof NO, NO2, O3, OH, HO2, H2CO, actinic flux, and temperature during the 1999‘Southern Oxidant Study’ (June 15 to July 15) at Cornelia Fort Airpark, Nashville,Tennessee. Averaged PO3 is plotted as a function of the NO mixing ratio. The datawere placed into three PHOx bins: high (0.5 < PHOx < 0.7 ppt s−1, circles), moderate(0.2 < PHOx < 0.3 ppt s−1, squares), and low (0.03 < PHOx < 0.07 ppt s−1, trian-gles), and then averaged as a function of NO. All three PHOx regimes demonstratethe expected generic dependence on NO. PO3 increases linearly with NO for low NO(<600 ppt NO), and then PO3 becomes independent of NO for high NO (>600 pptNO). The crossover point between NOX -limited and NOX -saturated O3 productionoccurs at different levels of NO in the three PHOx regimes (from Thornton et al.,2002, Copyright by American Geophysical Union (AGU), reproduced by permissionof AGU).
2.4 Radical Processes in the Atmosphere
Free radicals are the driving force for most chemical processes in the atmo-sphere. Since the pioneering work of Weinstock (1969) and Levy (1971), pho-tochemically generated HOX radicals (hydrogen radicals, OH + HO2) havebeen recognised to play a key role in tropospheric chemistry. In particular,hydrogen radicals:
• initiate the degradation and thus the removal of most oxidisable tracegases emitted into the atmosphere
• give rise to the formation of strongly oxidising agents (mostly in the tro-posphere), such as ozone or hydrogen peroxide
• catalytically destroy stratospheric ozone (see Sect. 2.10)
28 2 Atmospheric Chemistry
0:00:00 0:00:00 0:00:00 0:00:00 0:00:0014.08.13.08.12.08.11.08.
0
50
100
150
200
HeilbronnWelzheimer Wald
μg m–3
Fig. 2.8. Diurnal variation of ozone levels in different types of air masses duringAugust 11–14, 2000: Forest (Weltzheimer Wald, Germany) and city of Heilbronn(southern Germany).Source: Landesanstalt fur Umweltschutz Baden-Wurttemberg and UMEG
• are difficult to remove once they are generated, since radical-molecule re-actions tend to regenerate radicals.
Today we have an enormous amount of direct and indirect evidence of thepresence of HOX radicals (see, for example, Ehhalt et al., 1991; Wennberget al. 1998; Platt et al., 2002), and the importance of HOX for atmosphericchemistry can be assumed to be proved beyond reasonable doubt. Neverthe-less, the possible role of other radicals, beginning with the (historical) idea ofthe impact of oxygen atoms O(3P) or excited oxygen molecules O2(1Δ) hasbeen the topic of past and current investigations. In particular, the nitrateradical, NO3, (see Sect. 2.5.2) and the halogen atoms and halogen oxide rad-icals BrO, IO, and ClO (Sect. 2.8) can make a considerable contribution tothe oxidising capacity of the troposphere. For instance, reaction with NO3 orBrO can be an important sink of DMS in marine environments. In addition,night-time reactions of nitrate radicals with organic species and NOx playan important role for the removal of these species. NO3 chemistry can alsobe a source of peroxy radicals (such as HO2 or CH3O2), and even OH radi-cals (Sect. 2.5.2). Table 2.6 shows an overview of the most important radicalspecies in the troposphere and their significance for atmospheric chemistry.The details of the chemistry of NO3 and halogen oxides will be discussed inthe following sections. Here, we will concentrate on the tropospheric chemistryof hydroxyl radicals.
2.4 Radical Processes in the Atmosphere 29
Table 2.6. Free radical cycles pertinent to tropospheric chemistry, and key processesinfluenced or driven by reaction of those radicals
Species Significance
HOX cycle OH Degradation of most volatile organic com-pounds (VOC)Key intermediate in O3 formationNOX ⇒ NOY conversion
HO2 Intermediate in O3 formationIntermediate in H2O2 formation
RO2 Intermediate in ROOR’ formationAldehyde precursorPAN precursorIntermediate in O3 formation
NO3 cycle NO3 Degradation of certain VOC (olefins, aromat-ics, DMS, etc.)NOX ⇒ NOY conversion (via N2O5 or DMS-reactions)RO2 precursor (night-time radical formation)
XOX cycle XO (X = Cl, Br, I) Catalytic O3 destruction (cause of ‘PolarTrop. Ozone Hole’)Degradation of DMS (BrO)Change of the NO2/NO (Leighton) ratio
X Degradation of (most) VOC (Cl)Initiates O3 formationRO2 precursorInitiates particle formation (IOX)
2.4.1 Sources of Hydrogen Radicals (OH and HO2)
Hydroxyl radicals are probably the most important free radicals in the at-mosphere. The degradation of most (but not all) oxidisable trace gases (suchas HC or CO) is initiated by OH reaction (cf. Platt, 1999). OH radicals aretherefore sometimes called the ‘cleansing agent’ of the atmosphere. Figure 2.9shows a simplified outline of the HOX (OH + HO2) and RO2 cycle.
Several primary OH production mechanisms are known:(a) Globally, the most important process forming OH is initiated by UV
photolysis of ozone to form electronically excited oxygen atoms:
O3 + hν(λ < 320 nm) → O(1D) + O2(1Δ) (R2.4a)
Recent research shows that photolysis of vibrationally excited ozone can con-siderably enhance the rate of reaction (R2.4a). In addition, some contributioncomes from the channel:
O3 + hν(λ < 420 nm) → O(1D) + O2(3Σ) (R2.4b)
30 2 Atmospheric Chemistry
NO(O2)
CH3O2
CH2O
OH(O2)hν(O2)
hν
hν
O3(O2)
OH(O2)
CH4(NMHC)
P(OH)
‘Primary’OH
source
‘Final’ OH Sink
HCHO
HONO
Olefins
OH
OH
NONO2
H2O
HO2
H2O2HNO3
H2O
CO(O2)
H2(O2)
CH2O(O2)
O3
O3
O3
O3
O3
CH3O2(CnH2n+1– O2)
[HOX] α √P(OH)[HOX] α P(OH)
Fig. 2.9. A simplified outline of the HOX(OH+HO2) and RO2 cycles in the tropo-sphere. The very reactive OH attacks most oxidiseable species (e.g. CO, hydrocar-bons). Typical OH lifetimes are therefore below 1 s. Consequently, noontime concen-trations only reach 106–107 cm−3 (0.04–0.4 ppt). These reactions usually lead to theformation of HO2 or RO2, which, in turn, can be reconverted to OH by reaction withNO or O3, thus preserving HOX . Alternatively HOX is destroyed by self-reactionof HO2, producing H2O2. The only other ‘final’ HOX , sink of importance is theOH + NO2 recombination. In a very simplified picture, one can imagine the HOX
reservoir to be fed by R2.4 followed by R2.6 (rate of OH production = POH) anddrained by either HO2 self-reaction (if NOX is low) or OH + NO2 recombination(at high NOX). All other reactions (inside the box) only interconvert OH and HO2
(partly via RO2). Note that [HOX ] ∝√
POH at low NOX and [HOX ] ∝ POH at highNOX (i.e. NOX > 1 ppb)
(see e.g. Talukdar et al., 1998). The excited oxygen atoms O(1D) from (R2.4)can be deactivated by collision with N2, O2, or H2O (i.e. any molecule ‘M’):
O(1D) + M → O(3P) + M (R2.5)
A certain fraction X of the excited oxygen atoms, however, will react accordingto:
2.4 Radical Processes in the Atmosphere 31
O(1D) + H2O → OH + OH (R2.6)Assuming k2.5a and k2.5b denote the rate constants of the reaction of O(1D)with N2 or O2, respectively, and k2.6 the rate constant of reaction (R2.6), weobtain for X:
X =k3 · [H2O]
k2.5a · [N2] + k2.5b · [O2] + k2.6 · [H2O](2.16)
(Chemical symbols in square brackets denote the concentration of the partic-ular species in the atmosphere.) In the lower troposphere, X is typically in therange of 0.05–0.1. At very high relative humidity, up to 20% can be reached.The production rate of OH radicals, P (OH), is accordingly calculated fromthe O3-concentration and the photolysis frequency [see (2.18) below] for theabove process (R2.4a):
d
dt
[O(1D)
]= [O3] · J1 (2.17)
P (OH) =d
dt[OH] = 2 · X · [O3] · J1 (2.18)
(b) Another important source of OH is the reaction of HO2 radicals withNO:
HO2 + NO → OH + NO2 (2.19)
An analogous reaction of HO2 with ozone (R2.18) also yields OH. HO2
radicals are largely formed by OH reactions. The consequences of this ‘radicalrecycling’ will be analysed below. In addition, the photolysis of aldehydesforms HO2. This process is particularly significant in urban areas, where theoxidation of VOC and direct emissions lead to high aldehyde levels. A numberof non-photochemical mechanisms also produce HO2 radicals. Among these,reactions of NO3 radicals with certain organic species, such as olefins, phenols,or DMS, are particularly important (e.g. Platt et al. 1990, 2002; Geyer et al.,2003a; Geyer and Stutz, 2004; see Sect. 2.5.2).
(c) A further OH source is the photolysis of HONO (nitrous acid, seeSect. 2.5.3):
HONO + hν → OH + NO (R2.8)
This OH source is only of importance in polluted air, where HONO isformed by heterogeneous processes, i.e. chemical reactions on surfaces e.g. ofbuilding walls or aerosol particles (see, e.g. Platt 1986, Alicke et al. 2002, 2003Sect. 2.5).
(d) The ozonolysis of olefins also produces OH radicals (Atkinson et al.,1992; Atkinson and Aschmann, 1993; Paulson et al., 1997, 1999):
32 2 Atmospheric Chemistry
Olefin + O3 → OH + products (R2.9)
An ozone molecule adds across the C=C double bond to form a primaryozonide, which decomposes to form vibrationally excited carbonyl oxide andcarbonyl products. The subsequent chemistry in the gas phase is still notcompletely understood. For a recent publication, see e.g. (Finlayson-Pitts andPitts, 2000; Paulson et al., 1999). The OH production yields of these reactionsvary from 7% to 100%, depending on the structure and size of the alkene(Paulson et al., 1997; 1999). Note that this source (though usually weak whencompared with others) is independent of the light, and thus also providesradicals at night.
(e) Finally, photolysis of hydrogen peroxide produces OH radicals:
H2O2 + hν → OH + OH (R2.10)
In contrast to processes (b) and (c), the latter source would only be relevantin air masses with very low NOX pollution, since only under these conditionscan significant amounts of H2O2 be formed.
2.4.2 Temporal Variation of the HOX Source Strength
The photolysis frequency J2.4, and thus the strength of the main OH source[P(OH)], is given by the expression (e.g. Junkermann et al., 1989): whereI(λ) denotes the solar intensity (actinic flux), σ(λ) the ozone absorption cross-section, and Φ(λ) the quantum efficiency of reaction R2.4. Because the solarintensity varies with the time of day, cloudiness, and seasons, the OH sourcestrength is highly variable in the atmosphere. In addition, other environmen-tal variables, such as the water vapour concentration, influence the formationof OH radicals. Figure 2.10, for example, shows the diurnal variation of OHconcentrations. The OH levels follow the diurnal behaviour of the O3 pho-tolysis rate closely. The seasonal dependence of the OH source strength atmid-latitudes is shown in Fig. 2.11.
In polluted air, the diurnal variation of the ozone concentration (with amaximum in the early afternoon) has an additional influence on OH. Sincethe rate of R2.4 is proportional to the ozone concentration, P(OH) should alsobe proportional to [O3]. However, a higher O3 level will shift the NO/NO2
stationary state towards higher NO2 (see Sect. 2.5.1), and thus will reduce theOH concentration. The strength of other OH sources (e.g. HONO – photolysis,see Sect. 2.5.1) frequently peak in the morning.
2.4.3 Chemistry of Hydrogen Radicals (OH and HO2)
Hydroxyl radicals react with most oxidisable gases in the atmosphere, forinstance with carbon monoxide:
2.4 Radical Processes in the Atmosphere 33
00:00
CO
(pp
b)
NO
(pp
b)N
O2
(ppb
)
HO
2 (1
08 cm
–3)
OH
(106
cm–3
)
J (O
1 D) (1
0–5 s
–1 )
0
50
100
150
2000
1
2
3
0
1
2
3
4
5
06:00 12:00
Time (UTC)
18:00 00:00
0
1
2
3
4
5
0.0
0.5
1.0
1.5
0.0
0.5
1.0
1.5
2.0OH
NO
CO
HO2
NO2
J(O1D)
Fig. 2.10. Average diurnal OH, HO2, and J(O1D) levels as observed during theBERLIOZ field experiment near Berlin in summer 1998. The OH concentrationfollows the solar radiation closely, while the levels of HO2 are also influenced by theconcentrations of NO (from Holland et al., 2003, Copyright by American GeophysicalUnion (AGU), reproduced by permission of AGU).
OH + CO → CO2 + H (R2.11)
In this reaction, both CO2 and hydrogen atoms are produced. In the atmo-sphere, the hydrogen atoms immediately form hydro-peroxy radicals, HO2:
H + O2M−→ HO2 (R2.12)
As products of the degradation of VOCs, peroxy radicals are also produced.An example is methane degradation:
34 2 Atmospheric Chemistry
10
1 × 106
2 × 106
3 × 106
4 × 106
5 × 106
6 × 106
7 × 106
2 3 4 5 6
Month
OH
pro
duct
ion
rate
cm
–3 s
–1
7 8 9 10 11 12
Fig. 2.11. The seasonal variation of the average diurnal OH source strength undercloud-free conditions in mid-latitudes (Germany)
OH + CH4 → CH3 + H2O (R2.13)
CH3 + O2M−→ CH3O2 (R2.14)
Further important reactions of OH are the oxidation of molecular hydrogenand formaldehyde (and higher aldehydes):
OH + H2 → H2O + H (R2.15)
OH + CH2O → CHO + H2O (R2.16)
CHO + O2 → HO2 + CO (R2.16a)
In these reactions, the primary product is also HO2 (R2.15 followed by R2.12,or R2.16 followed by R2.16a, respectively)
Thus, in all of the above reactions, and in fact most OH reactions, freeradicals (here HO2 radicals, which are easily converted to OH) are regenerated.An important exception to this rule is the reaction of OH with oxides ofnitrogen:
OH + NO2M−→ HNO3 (R2.17)
The nitric acid produced in this reaction will (in the troposphere) be removedfrom the gas phase by various processes, such as dry deposition to the groundor rainout. Figure 2.9 shows the simplified reaction scheme of the hydrogenradicals HOX(OH + HO2), as outlined above.
2.4 Radical Processes in the Atmosphere 35
The balance between OH sources and sinks determines the levels of OH andHO2 and their respective lifetimes. Figure 2.12 shows the OH and HO2 budgetas a function of the NOX (NO + NO2) concentration. The OH concentrationshows a non-monotonous behaviour as a function of the NOX level: At verylow NOX (below e.g. 0.1 ppb) essentially each OH radical reacting with CO (orHC) is converted to HO2 (or an organic peroxy radical, RO2). These radicalsare, in turn, lost by radical–radical interaction. Thus, the OH concentrationremains relatively low. At medium NOX levels (about 1–2 ppb) most HO2 andRO2 radicals react with NO to reform OH. Therefore, OH levels are muchhigher than in the case of low NOX . At very high NOX levels (10 ppb orhigher) most of the OH radicals react with NO2 to form HNO3, and thereforeOH levels are low again.
As a consequence, OH levels show a maximum at around 1 ppb of NOX ,while HO2 levels drop monotonously with increasing NOX . This behaviour isillustrated in Fig. 2.13.
Although its concentration (in clean air) can be up to two orders of magni-tude larger than that of OH (see Fig. 2.13), the reactivity of the HO2 radicalis much lower than that of the OH radical. However, the HO2 radical is animportant OH reservoir, and it is also the precursor for atmospheric H2O2. Inaddition, there is most likely a direct role of HO2 in liquid-phase chemistry,for instance in cloud droplets (Chameides and Davis, 1982; Chameides, 1984;Lelieveld and Crutzen, 1990, 1991).
Only four HO2 reactions play a role in the troposphere. Their significancedepends on the levels of NOx:
(1)HO2 + NO → OH + NO2 (R2.7)
(2)HO2 + O3 → OH + 2O2 (R2.18)
(3)HO2 + HO2 → H2O2 + O2 (R2.19)
(4)HO2 + OH → H2O + O2 (R2.20)
The first two reactions reduce HO2 radicals to OH. The third and forth reac-tions can be regarded as ‘final sinks’ for the HO2 radical (and thus for HOX).While hydrogen peroxide can be photolysed to yield two OH radicals (R2.10),this is a slow process. Thus rainout and washout of H2O2 is much more likely,since it is highly water soluble. A further sink of H2O2 is reaction with OH:
H2O2 + OH → H2O + HO2 (R2.21)
Hydrogen Peroxide is an important oxidant in the liquid phase, e.g. in rain-drops, or cloud or haze droplets. Reactions of H2O2 in cloud or haze dropletscontribute significantly to the oxidation of SO2 (S(IV)) to sulphate (S(VI)).
36 2 Atmospheric Chemistry
OH
OH
4 × 106 cm–3
τ~0.9 s
14 × 106 cm–3
τ~0.6 s
OH
3 × 106 cm–3
τ~0.03 s
20%
80%
30%
70%
10%
20%
86%
96%
14%(Peroxides)
4%
17%
1%
2%
82%
90%
[NOX] = 0
[NOX] = 2 ppb
[NOX] = 100 ppb
O3 + hν
+ H2O
+ HO2
+ NO2
+ HO
NO
+ NO
NO
HNO
HO2
HO2
HO2
O3
H2O
HNO3
H2O2
(Peroxides)H2OH2OO3 + hν
+ H2O
O3 + hν
+ H2O
(+ O2)CH2O + hν
Fig. 2.12. The relative strength of OH sources and sinks, and the resulting OHconcentration as a function of the NOX (NO + NO2) level. Here, three cases areshown: (a) At very low NOX , essentially each OH radical reacting with CO (orhydrocarbons) is converted to HO2 (or an organic peroxy radical, RO2). These rad-icals are, in turn, lost by radical–radical interaction, and thus the OH concentrationremains relatively low. (b) At medium NOX levels (about 1 ppb), most HO2 andRO2 radicals react with NO to reform OH. Therefore OH levels are much higherthan in case (a). (c) At very high NOX levels, most of the OH radicals react withNO2 to form HNO3, and therefore OH levels are low again
Hydrogen peroxide is essentially only formed via reaction R2.19 (i.e. HO2 re-combination). A small contribution is assumed to come from the reaction ofolefines with ozone (Becker et al., 1990). Assuming reaction R2.19 to be thedominating source, high rates of H2O2 production (and thus usually also high
2.4 Radical Processes in the Atmosphere 37
100
101
0.10.01
101
0.10.01
0
2
6
4
0
2
6
8
4
5001000
100
500
NOX (ppbv)
NOX (ppbv)CO (p
pbv)
CO (ppbv)
1000
HO
2 (1
08 cm
–3)
OH
(106
cm–3
)
Fig. 2.13. Concentrations of OH (upper panel, units of 106 molecules cm−3) andHO2 (lower panel, units of 108 molecules cm−3) as a function of the NOX (NO+NO2)mixing ratio for typical conditions (37.6 ppb ozone, 88.7 ppb CO, 0.92 ppb CH2O,J(O3 − O1D) = 3 × 10−5 s−1, J(NO2) = 9.1 × 10−3 s−1). The HO2 concentration,and therefore H2O2 production rate, are highest in unpolluted air at NOX levelsbelow a few 100 ppt (from Enhalt, 1999)
38 2 Atmospheric Chemistry
levels) are to be expected under conditions of high HO2. Thus, in air masseswith high humidity and ozone, and low NO, the highest H2O2 levels are to beexpected.
2.5 Oxides of Nitrogen in the Atmosphere
The oxides of nitrogen, NO and NO2 (NOX), are key species in atmosphericchemistry. They regulate many trace gas cycles and influence the degradationof most pollutants in clean air, as well as in polluted regions:
NOX concentration has a strong influence on the atmospheric level ofhydroxyl radicals (OH, see Sect. 2.4.1), which, in turn, are responsible for theoxidation processes of most trace gases. In addition, NOX is a catalyst fortropospheric ozone production (see Sect. 2.3). Oxides of nitrogen (or acidsformed from them) can also react with VOC degradation products to formorganic nitrates or nitrites [e.g. peroxy acetyl nitrate (PAN) or methyl nitrite],as well as nitrosamines (Platt et al., 1980b). These species can be much moredetrimental to human health than the primary oxides of nitrogen. Finally,the thermodynamically most stable and ultimate degradation product of allatmospheric oxides of nitrogen, nitric acid, is (besides sulphuric acid) the mainacidic component in ‘acid rain’.
An overview of the most important oxidised nitrogen species in the atmo-sphere is given in Fig. 2.14. The main reaction pathways between the variousspecies are indicated by arrows. Oxides of nitrogen are primarily emitted inthe form of NO (plus some NO2) and N2O. While N2O is a very inert speciesand therefore plays no role for the chemical processes in the troposphere, NOreacts rapidly with natural ozone to form NO2. On the other hand, NO2 willbe destroyed (photolysed) by sunlight during the daytime, and therefore a sta-tionary state between these two NOX species will be established. NO2 thenfurther reacts with OH radicals or NO3, ultimately forming nitric acid (ornitrate aerosol).
The ‘classic’ role of oxides of nitrogen in atmospheric chemistry (seeSect. 2.5.1) has been known since the 1980s. Since then, the importance ofthis class of species has been underlined by newly obtained results, largelymade possible by improved measurement techniques:
• The confirmation of an additional NOX degradation pathway via hetero-geneous and homogeneous hydrolysis of N2O5 at night (Heintz et al., 1996;Mentel et al., 1996; Wahner et al., 1998a; Geyer et al., 2001c; Brown et al.,2003; Stutz et al., 2004).
• The non-photochemical production of peroxy radicals (HO2 and RO2) viaNO3–VOC reactions as a further consequence of elevated NOX levels inthe atmosphere (Platt et al., 1990; LeBras et al., 1993; Mihelcic et al.,1993; Geyer et al., 1999; Geyer and Stutz, 2004).
2.5 Oxides of Nitrogen in the Atmosphere 39
Fig. 2.14. Simplified overview of the NOX reaction scheme in the atmosphere.Arrows indicate main reaction pathways
• The establishment of HONO photolysis as another significant OH sourcein polluted air (Perner and Platt, 1979; Harris et al., 1982; Platt, 1986;Harrison et al., 1996; Alicke et al., 2002, 2003).
• Finally, an important role of oxidised nitrogen species is postulated in themechanisms releasing RHS from sea salt as, for example, observed in thepolar BL (Finlayson-Pitts and Johnson, 1988; Finlayson-Pitts et al., 1989,1990).
The improved knowledge of the cycle of oxidised nitrogen species also hasgreat practical importance: The exact knowledge of the NOX lifetime allowsthe estimation of NOX levels and the health effects of atmospheric NOX , theamount of long-range transport (and thus, for instance, the amount of acidrain in non-polluted areas), and finally the NOX level in clean air (and thusthe amount of ozone formation there).
2.5.1 Classical Chemistry of Oxides of Nitrogen in the Atmosphere
In this section, we briefly summarise the central aspects of the classical chem-istry of oxides of nitrogen in the atmosphere (see also Fig. 2.14 and Table 2.3).
40 2 Atmospheric Chemistry
Extensions of our image of the chemical processes related to oxidised nitrogenspecies are discussed in Sect. 2.5.2.
The transformation pathways between the atmospheric nitrogen reservoirare closely coupled to the reactions of ozone (as discussed in Sect. 2.3) and tothe cycles of hydrogen radicals (OH and HO2). An important question in thiscontext is: Are NOX reactions sources or sinks for ozone? The simple systemdescribed by reactions R2.1–R2.3 (Sect. 2.3) neither destroys nor producesozone. However, in the atmosphere there are additional reactions oxidisingNO to NO2 without consuming ozone. Thus, the NOX level ultimately deter-mines whether there is net ozone formation or destruction. In the unpolluted(background) atmosphere, the processes oxidising NO to NO2 without con-suming O3 (and thus producing O3 via reactions R2.1 and R2.2) are reactionsof NO with peroxy radicals (HO2, CH3O2, and higher RO2). In the simplecase of HO2, we have:
NO + HO2 → NO2 + OH (R2.7)
The HO2 radicals are produced, for example, by reaction of OH radicalswith CO (R2.11, followed by R2.12). At the same time, the reactions R2.7,R2.11, and R2.12 facilitate another stationary state in the atmospheric con-centration ratio of the hydrogen radicals OH and HO2, (HOX). Since ozonealso reacts with HO2 radicals (but not with other peroxy radicals), the HOX
system also destroys ozone:
O3 + HO2 → OH + 2O2 (R2.18)
Whether the above-described interplay of the NOX and HOX systemsleads to net ozone formation (as shown in Fig. 2.7) or O3 destruction dependson the relative rates of the reactions R2.7 and R2.18, and thus on the NOconcentration. At a NO concentration around 1/3000 of the O3 concentration(i.e. [NO] ≈ 0.01 ppb at 30 ppb O3), the rates of both reaction pathways areequal. In other words, ozone formation equals ozone destruction. At higherNOX (and thus NO) levels net O3 formation occurs, while at lower NO levelsO3 is photochemically destroyed. In regions influenced by industrial activities,where NO levels can reach many ppb, O3 formation by far exceeds destruction.This can also be seen from the observation that ozone levels in the (more in-dustrialised) northern hemisphere exceed those in the southern hemisphere byabout 50%. Figure 2.7 shows model calculated and observed ozone formationrates as a function of the NO level.
Oxides of nitrogen influence the concentrations of OH and HO2 radicals inthe atmosphere by several mechanisms. The most important removal processfor OH radicals is their recombination with NO2. This reaction (R2.6) leadsto final removal of OH from the radical chain. Consequently, the OH concen-tration is reduced by high levels of NOX (and thus NO2). On the other hand,the reaction of NO with HO2 radicals reduces the stationary state [HO2]/[OH]ratio (reactions R2.7, R2.11, and R2.12), and consequently leads to higher OHradical concentrations. This relationship is illustrated in Figs. 2.12 and 2.13.
2.5 Oxides of Nitrogen in the Atmosphere 41
The crucial step initiating the eventual removal of nitrogen oxides from theatmosphere is their conversion to nitric acid (HNO3), which is subsequentlyremoved from the atmosphere by dry and wet deposition (see also Table 2.3).The direct (dry) deposition of nitrogen oxides to the ground is of minor im-portance in comparison (<10%). On the other hand, re-conversion of nitricacid to NOX is negligible in the troposphere. Nitric acid is formed in reactionsof NO2:
• Directly by reaction with OH (R2.17).• Indirectly by reaction of NO2 with O3, leading to NO3 (R2.22), which can
either abstract a hydrogen atom (e.g. from DMS) or further react withNO2 to form N2O5, (R2.26) the anhydride of nitric acid, which in turn isconverted to HNO3 by reaction with gas-phase or liquid water (R2.27).
2.5.2 Tropospheric Chemistry of Nitrate Radicals, NO3
At night, when levels of OH radicals are low, other oxidants play an importantrole in the troposphere. Among those, the NO3 radical plays an importantrole for the budget of NOX and for the concentrations of certain organiccompounds. Nitrate radicals are formed by reaction of NO2 with ozone (seealso Fig. 2.14)
NO2 + O3 → NO3 + O2 (R2.22)
which constitutes the only relevant source of NO3 in the troposphere. Therate of NO3 production, PNO3, can be calculated from the concentration ofNO2 and O3:
d [NO3]dt
= [NO2] · [O3] · k1 = PNO3 (2.20)
During the daytime, NO3 radicals are effectively destroyed by photolysis:
NO3 + hν → NO + O2 (J2.23a)
→ NO2 + O (J2.23b)
(photolytic lifetime is about 5 s), and rapid reaction with NO (k2.24 = 2.6 ×10−11 cm3s−1 at 293K):
NO3 + NO → NO2 + NO2 (R2.24)
At night, when photolysis is unimportant and NO levels are usually low, theNO3 loss rate is much smaller and thus its concentration is larger. Loss ofNO3 due to gas-phase reaction with nitrogen dioxide or the NO3 self-reactionis slow. Similarly, the unimolecular decomposition of NO3 is probably of minorimportance in the atmosphere (Johnston et al., 1986; Russel et al., 1986). Gas-phase reaction of NO3 with water vapour is endothermic.
Main destruction mechanisms are thus reaction with organic species (seeTable 2.7) or heterogeneous loss (i.e. reaction at the ground or at the surfaceof aerosol or cloud particles) of either NO3 or N2O5.
42 2 Atmospheric Chemistry
Table 2.7. Reaction of OH, NO3, O3, and Cl with VOC (at 25◦C) (from Atkinson,1994)
Species Reaction rate constant 1015 cm3 molec.−1 s−1
With OH With NO3 With Ozone With Cl
AlkanesMethane 6.14 <0.001 <0.00001 100Ethane 254 0.0014 <0.00001 59000Propane 1120 0.017 <0.00001 137000n-Butane 2190 0.046 <0.00001 205000n-Pentane 4000 0.04 <0.00001 280000
AlkenesEthene 8520 0.21 0.0017 99000Propene 26300 9.4 0.013 2300001-Butene 31400 12 0.011 2200002-Methyl-2-Butene 68900 9300 0.4232,3-Dimethyl-2-Butene 110000 57000 1.16
AlkynesEthyne 5100 0.0082 <0.00001
AromaticsBenzene 1230 <0.03 <0.00001Toluene 5960 0.068 <0.00001Phenol 26300 3780o-Cresol 42000 13700 0.00026m-Cresol 64000 9740 0.00019p-Cresol 47000 10700 0.00047Benzaldehyde 12900 2.6 No dataIsoprene 101000 678 0.013 460000
Terpenesα - Pinene 53700 6160 0.0866 480000β - Pinene 78900 2510 0.0153-Carene 88000 9100 ≈0.037Limonene 171000 12200 ≈0.2Sabinene 117000 10000 ≈0.086α - Terpinene 363000 180000 8.47Myrcene 215000 11000 0.47
AldehydesFormaldehyde 9370 0.6 <0.00001 73000Acetaldehyde 15800 2.7 <0.00001 79000
NO3 + VOC Reactions
NO3 reacts with a variety of organic compounds in the atmosphere. In par-ticular, reactions with biogenic VOC, such as isoprene and terpenes, as wellas the reactions with DMS, are of significance.
2.5 Oxides of Nitrogen in the Atmosphere 43
NO3 reactions with alkanes and aldehydes proceeds via H abstraction (Atkin-son, 1990):
NO3 + RH → R + HNO3 (R2.25)
The alkyl radicals produced in reaction R2.25 are transformed to peroxyradicals (RO2) by reaction with O2.
Reactions with alkenes (including isoprene and terpenes) proceed via NO3
addition to a C = C double bond. Further reaction steps are O2 addition,followed by either reaction with other peroxy radicals or NO2. Reasonablystable end products are organic nitrates and peroxy radicals (Atkinson, 1990).
Reactionswith aromatic compounds are also important.While reactionwithbenzene is quite slow (rate constant on the order of 10−17 cm3molec.−1s−1),hydroxy-substitutedbenzenes reactvery fast,with reaction rate constantson theorder of 10−12 cm3 molec.−1 s−1 (furane and phenol) to 10−11 cm3 molec.−1 s−1
(cresols).Finally, NO3 radicals react rapidly with DMS (reaction rate constant
10−12 cm3 s−1). This can be significant for the budgets of both DMS andNO3, in particular in the marine atmosphere. The exact reaction mechanismis still subject to investigations. Recent investigations show that the primarystep is probably H abstraction. Thus, nitric acid and peroxy radicals shouldbe the dominant products (Butkovskaya and LeBras, 1994; Platt and LeBras,1997).
Heterogeneous Loss of NO3 and N2O5
In the atmosphere, NO3 and NO2 are in a chemical equilibrium with N2O5:
NO2 + NO3M←→ N2O5 (R2.26)
The equilibrium concentration of N2O5 is given by:
[N2O5] = K · [NO2] · [NO3] (2.21)
Under ‘moderately polluted’ conditions (NO2 mixing ratio of about 1 ppb)typical for rural areas in industrialised countries, NO3 and N2O5 concentra-tions are roughly equal (293K) (e.g. Wangberg et al., 1997). While homoge-neous reactions of N2O5 with water vapour and other atmospheric trace gasesare believed to be extremely slow (Mentel et al., 1996), heterogeneous reactionwith water contained in aerosol or cloud particles will convert N2O5 to nitricacid:
N2O5 + H2Ohet−→ HNO3 (R2.27)
Accommodation coefficients γ for N2O5 and NO3 on water (droplet) surfacesare in the range of 0.04–0.1 (the higher values applying to water containing
44 2 Atmospheric Chemistry
H2SO4) (Van Doren et al., 1990; Mozurkewich and Calvert, 1988; Mentelet al., 1996; Wahner et al., 1998; Hallquist et al., 2003), and >0.0025 (Thomaset al., 1993), respectively. If reaction R2.27 leads to accumulation of nitratein the liquid phase, the accommodation coefficients can be considerably (byabout one order of magnitude) reduced. For instance, Wahner et al. (1998)find γ in the range of 0.002–0.023 at relative humidities of 48–88% for theaccommodation of N2O5 on sodium nitrate aerosol.
NO3 Lifetime
From measured concentrations of O3, NO2, and NO3, the atmospheric lifetimeof NO3 (τNO3, see Sect. 2.1.3), limited by the combination of any first-orderloss process, can be calculated (Heintz et al., 1996; Geyer et al., 2001a,b,c,2002; Brown et al., 2003; Geyer and Stutz, 2004):
NO3 + Z → products (R2.28)
Assuming pseudo stationary state conditions with respect to NO3, τNO3 be-comes
τNO3 =[NO3]PNO3
=[NO3]
[NO2] · [O3] · k1(2.22)
Since k2.22 is known from laboratory measurements, and the concentrationsof O3, NO2, and NO3 can be simultaneously measured (for instance, by theDOAS technique), τNO3 can readily be calculated for various atmosphericconditions. It must be emphasised that the steady-state assumption is anapproximation, which is quite good for most atmospheric situations. The NO3
lifetime can also be limited by any irreversible loss of N2O5 (for instancek2.27 > 0), because both species are in a rapidly established thermodynamicequilibrium (R2.28, the time constant to reach equilibrium, is on the order of1 min at ambient temperatures). Under conditions where both NO3 and N2O5
are lost, the observed NO3 lifetime τNO3 is given by:
τNO3 =1
1τ ′NO3
+[NO2]
K· 1τN2O5
(2.23)
Of course, in the case of negligible loss of N2O5(τN2O5 → ∞), τ ′NO3 = τNO3.
For conditions where direct loss of NO3 molecules can be neglected (τ ′NO3 →
∞), (2.23) reduces to:
τNO3 =K
[NO2]· τN2O5 (2.24)
The simple stationary state assumption predicts an observed NO3 lifetimeτNO3 inversely proportional to the NO2 concentration in cases where only
2.5 Oxides of Nitrogen in the Atmosphere 45
N2O5 is lost. Since NO3 formation is proportional to [NO2] (R2.20), [NO3]should become independent of the NO2 concentration under such conditions.
Formation of Hydrogen Radicals by NO3 + VOC Reactions
Since most of the above reactions of NO3 with organic species appear to leadto the formation of peroxy radicals, an interesting consequence of the night-time reaction of NO3 with organic species is a considerable contribution tothe HOX budget (Platt et al., 1990; Mihelcic et al., 1993; Carslaw et al., 1997;Geyer et al., 2001; Geyer and Stutz, 2004).
In circumstances where NO levels are very low, peroxy radicals undergoreactions with other RO2 radicals, O3, or NO3:
RO2 + NO3 → RO + NO2 + O2 (R2.29)
In urban areas, where ubiquitous NO emissions and vertical transport can leadto an active mixing of NO3, N2O5, and NO, the likely fate of RO2 radicals isreaction with NO (Geyer and Stutz, 2004):
RO2 + NO → RO + NO2 (R2.30)
This gives rise to a radical chemistry which is similar to that during theday. However, the region in which NO3 and NO can coexist is often shallowand it moves vertically throughout depending on the NO emission strengthand vertical stability (Geyer and Stutz, 2004).
2.5.3 Nitrous Acid, HONO in the Atmosphere
The photolysis of nitrous acid, HONO, is an important source for OH radi-cals in the polluted urban atmosphere (Perner and Platt, 1979; Platt et al.,1980b; Kessler et al., 1981; Harris et al., 1982; Alicke et al., 2002, 2003; Zhouet al., 2002; Aumont et al., 2003; Ren et al., 2003; Kleffmann et al., 2005),and the polar BL (Li, 1994; Zhou et al., 2001). In addition, HONO is toxic(Beckett et al., 1995) and its chemistry leads to the formation of carcinogenicnitrosamines, (Shapley, 1976; Famy and Famy, 1976; Pitts, 1983). HONO for-mation at chamber walls is a candidate for an intrinsic ‘chamber dependent’radical formation, which is observed in most smog chamber experiments (e.g.Gleason and Dunker, 1989; Rohrer et al., 2004).
The photolysis of HONO, which absorbs in the wavelength range from300 to 405 nm, leads to a formation of an OH radical and NO (Stockwell andCalvert, 1978; Stutz et al., 2000):
HONO + hν → OH + NO (300 nm < λ < 405 nm) (R2.8)
The photolysis of HONO begins early in the morning, when HONO concentra-tions are typically highest. The combination of elevated HONO concentrations
46 2 Atmospheric Chemistry
at sunrise and fast photolysis results in a peak in the production of OH whichsurpasses other HOX sources, such as O3 and HCHO photolysis (Alicke et al.,2002).
Nitrous acid was first identified and quantified in the atmosphere by DOAS(Perner and Platt, 1979; Platt and Perner, 1980). Typical HONO mixing ratiosare in the range of 0 – 15 ppb (see Calvert et al., 1994; Lammel and Cape,1996; or Alicke, 2000 for reviews). Generally, nitrous acid concentrations scalewith the NO2 concentration, and therefore approximately with the degree ofpollution. The diurnal variation of the mixing ratio of HONO is dominatedby its photolysis, which leads to lower levels during the day. During the nightconcentrations increase, often reaching a constant value in the early morning.The amount of HONO formed is often correlated with the stability of thenocturnal BL, as inferred for example from radon data (Febo et al., 1996).
Several studies have discussed the importance of HONO photolysis as anOH source (e.g. Platt et al., 1980b; Harrison et al., 1996; Stutz et al., 2002;Alicke et al., 2002, 2003). Alicke et al. (2002) showed OH production rates ofup to 3 × 107 molecules cm−3 s−1 in Milan, Italy. Other observations confirmthese results (Harris et al., 1982; Sjodin, 1988). Model studies (e.g. Harriset al., 1982) show that HONO photolyis can lead to an increase in the maxi-mum daytime ozone concentration of up to 55%, if 4% of the initial NOX ispresent as HONO (Los Angeles case, initial NOx: 0.24 ppm), compared with acalculation without nitrous acid. Jenkin et al. (1988) showed HONO photol-ysis can account for a fivefold increase in OH at 6:00 AM, a 14% increase inOH present at the daily maximum (noon), and a 16% increase in net photo-chemical ozone production. In addition, the formation of ozone begins earlierin the day due to the HONO photolysis in the morning.
HONO is formed through various chemical mechanisms in the atmosphere.The only important gas-phase reaction forming nitrous acid is 2.31 (Stuhl andNiki, 1972; Pagsberg et al., 1997; Zabarnick, 1993; Nguyen et al., 1998):
OH + NO(+M) → HONO(+M) (R2.31)
Gas-phase nitrous acid producing pathways, such as the reaction of NO2 withHO2, have been reported to be of minor importance (e.g. Howard, 1977; Tyn-dall et al., 1995). The bulk of urban HONO is suspected to be formed fromeither of the mechanisms summarised in reaction R2.32 or R2.33, involvingonly NOX and water vapour (Perner and Platt, 1979; Kessler and Platt, 1984;Platt, 1986).
NO2 + NO + H2O ⇔ 2HONO (R2.32)2NO2 + H2O ⇔ HONO + HNO3 (R2.33)
These reactions (and their reverse reactions) appear to proceed hetero-geneously on surfaces (Kessler, 1984; Kessler et al., 1981; Sakamaki et al.,1983; Lammel and Perner, 1988; Notholt et al., 1991, 1992; Junkermann andIbusuki, 1992; Ammann et al., 1998; Calvert et al., 1994; Longfellow et al.,
2.6 Tropospheric Chemistry of VOCs 47
1998; Goodman et al., 1999; Kalberer et al., 1999; Veitel, 2002; Veitel et al.,2002). A number of laboratory studies (Sakamaki et al., 1983; Pitts et al.,1984b; Jenkin et al., 1988; Svensson et al., 1987; Kleffmann et al., 1998) haveshown reaction R2.32 to be insignificant. Several field observations, where thepresence of high ozone at night excluded NO, or low NO was documented,confirm this result (e.g. Kessler and Platt, 1984; Harrison and Kitto, 1994).
The exact mechanism of the heterogeneous formation of HONO sum-marised in reaction R2.33 is unknown, but several studies (Svensson et al.,1987; Jenkin et al., 1988; Kleffmann et al., 1998; Finlayson-Pitts et al., 2003)have shown that it is first order in NO2 and water. Neither the reaction rateconstants nor the nature of the surface is known, which makes calculationof the OH production by models difficult. In addition, reactions on ‘urbansurfaces’ like organic aerosols or soot aerosols were suggested (e.g. Gutzwilleret al., 2002; Broske et al., 2003). In fact reactions of NO2 on asphalt or rooftile surfaces appear to be sufficiently rapid to explain observations (Trick,2004).
The most important gas-phase chemical removal process for HONO, be-sides photoloysis, is reaction R2.34. A few percent of HONO is expected tobe destroyed by OH radicals:
OH + HONO → .H2O + NO2 (R2.34)
Several studies have been carried out to estimate the strength of directemission of nitrous acid from combustion processes (Kessler and Platt, 1984;Pitts et al., 1984a; Kirchstetter et al., 1996; Ackermann, 2000; Winer andBiermann, 1994). Up to 1% of the emitted NOX was found as nitrous acid,making this source important, especially in heavily polluted areas with highamounts of traffic. (Kirchstetter et al., 1996) report a HONO/NOX emis-sion ratio of 0.35% for a north-American car fleet, while (Ackermann, 2000;Kurtenbach et al., 2002) found 0.65% of NOX emitted as HONO in a traffictunnel in Germany.
2.6 Tropospheric Chemistry of VOCs
Organic compounds are an extremely large class of chemicals that play asignificant role in the atmosphere. As discussed in Sect. 2.3, they provide thefuel for ozone formation. In addition, chemical reactions with organic carbonscan lead to the formation of particles, so called secondary organic aerosol.They also impact radical budgets (see Sect. 2.4) and thus influence radicalchemistry.
The chemistry of volatile organic carbon compounds (VOCs) is a complextopic which, by far, exceeds the scope of this book. Here, we will only brieflydiscuss typical levels of organics in the atmosphere, and in particular in ur-ban areas, and their initial reactions with various radical species. VOCs canbe subdivided into a number of different classes encompassing hydrocarbons(HC) and oxidised species (Table 2.7):
48 2 Atmospheric Chemistry
• alkanes, with the overall formula CnH2n+2
• alkenes (or olefins), with the overall formula CnH2n (mono-olefins)• alkynes, with the overall formula CnH2n−2
• aromatics, which contain one or more benzene rings (C6H6)• isoprene and terpenes• oxidised species: aldehydes, ketones, alcohols, and organic acids
Members of each class often undergo similar chemistry. We will thereforediscuss their initial reactions steps in a general way.
Alkanes are molecules containing carbon and hydrogen molecules, whereasthe carbon molecules are held together by single molecular bonds. Conse-quently, the degradation of alkanes by OH proceeds by abstraction of onehydrogen atom, in analogy to the reaction of methane (R2.35 and R2.36):
OH + RH → R + H2O (R2.35)
where R denotes an alkyl radical (CnH2n+1), e.g. C2H5. The rate constantsfor some alkane–OH reactions are given in Table 2.7.
Analogous to CH3 (see Sect. 2.4.1), alkyl radicals immediately react withmolecular oxygen to form organic peroxy radicals:
R + O2M−→ RO2 (R2.36)
The fate of peroxy radicals in the atmosphere are self-reaction (formingperoxides), reaction with O3, and, mostly in polluted areas, reaction with NO:
RO2 + NO → RO + NO2 (R2.37)
Alkoxy radicals RO react with O2 to form aldehydes and HO2 radicals:
RO + O2 → R′CHO + HO2 (R2.38)
where R′ denotes an alkyl radical with one less C atom (Cn−1H2n−1).Initial reactions of alkanes with NO3 radicals and Cl atoms also proceed by
hydrogen abstraction, followed by reactions R2.36–R2.38 (Table 2.7). Ozoneonly reacts very slowly with alkanes and its ractions are unimportant in theatmosphere (Table 2.7).
Alkenes are HC which contain one or more carbon double bonds. Whilehydrogen abstraction can still occur, the predominant first step in alkene+OHreactions is the addition of the OH radical to the double bond, followed byaddition of O2 to the resulting peroxy radical:
OH + RCH2CH2M−→ RCH2CH2OH (R2.39)
RCH2CH2OH + O2M−→ O2RCH2CH2OH (R2.40)
Alkenes with aliphatic chains can also (though less likely) react by hy-drogen abstraction from the aliphatic chain. The rate constants for some of
2.6 Tropospheric Chemistry of VOCs 49
the more abundant alkenes with OH radicals are listed in Table 2.7. Similarto the alkane reaction chain, the O2RCH2CH2OH radical can react with NOforming a β-hydroxyalkoxy radical:
O2RCH2CH2OH + NO M−→ ORCH2CH2OH (R2.41)
The ORCH2CH2OH radical then further decays into hydroxyl or alkoxyradicals and aldehydes.
Reactions of alkenes with NO3 radicals proceed primarily through theaddition of NO3 to the double bond. The radical formed in the NO3 additionreacts further with O2 to form a β-nitratoalkyl peroxy radical, which canthen react with NO to form nitrates, hydroxyl radicals, and other products.Alkenes can also react with O3. Although these reactions are slower thanthose of OH, they can be important at night, when OH levels are low. Theirsignificance also stems from the fact that OH radicals are formed from thereaction of ozone and larger alkenes (Paulson et al., 1999). Cl atom reactions
Fig. 2.15. The initial steps in the OH initiated oxidation of benzene (1) Currentlyproposed loss processes of the aromatic-OH adduct (2) and the peroxy radical (3)(which are in rapidly established equilibrium) are shown that lead in part to theformation of phenol (4) Intermediate species are indicated by bold numbers. Similarschemes for the phenol forming pathways can also be drawn for the alkyl-substitutedaromatics (from R. Volkamer, 2001)
50 2 Atmospheric Chemistry
also proceed through Cl addition to the double bond. Details of the reactionchains for individual alkenes can, for example, be found in Finlayson-Pittsand Pitts (2000).
Alkynes contain carbon triple bonds. Their reactions proceed in similarways as those of alkenes. An important product of these reactions is glyoxal(CHOCHO).
For monocyclic aromatics the primary reaction pathway is also OH ad-dition. The resulting radical can either lose a hydrogen atom (through HO2
formation) and form a phenol, or the ring can open leading to rapid degrada-tion of the products (see Fig. 2.15) (Volkamer et al., 2002). In the latter case,glyoxal was found to be an important product (Volkamer et al., 2001).
Isoprene and terpenes are olefins with more than one double bond; theyare very reactive towards OH, NO3, O3, and Cl, with reaction rate constantstowards OH radicals approaching the collision limit. Because these moleculeshave a complex chemical structure, their reactions mechanisms are quite com-plex, and beyond the scope of this book. The reader is referred to a review byAtkinson and Arey (1998).
Aldehydes, such as formaldehyde and acetaldehyde, react with OH, NO3,and Cl by abstraction of the weakly bond aldehydic hydrogen (see for exampleR2.16). The radical formed in this reaction then reacts with O2, similar to thereactions described above. Ketones react by abstraction of the H atom fromthe alkyl chains, similar to the reactions of alkanes.
2.7 Tropospheric Chemistry of Sulphur Species
Sulphur compounds play an important role in the atmosphere, in clean air aswell as in polluted areas (e.g. Rotstayn and Lohmann, 2002). In particular,sulphur species with low vapour pressure (such as sulphuric acid or methanesulphonic acid) have a large influence on particle formation in the atmosphere.The legendary London Smog largely consisted of sulphuric acid aerosol. Themost important atmospheric sulphur species, their typical mixing ratios, res-idence times, and degradation pathways are summarised in Table 2.8. In thissection we will briefly discuss the chemistry of these species and the impactof sulphur chemistry on the climate.
2.7.1 Sulphur Dioxide – SO2
Sulphur dioxide is oxidised in the gas phase and, owing to its relatively highsolubility in water, by reactions in atmospheric liquid water, such as cloud,fog, or aerosol (see Fig. 2.16)
The most important gas-phase reaction is oxidation, which is initiated byOH addition:
2.7 Tropospheric Chemistry of Sulphur Species 51
Table 2.8. Sulphur compounds in the atmosphere
Species Chemicalformula
Typ. mixingratio ppb
Atmosphericlife-time
Degraded byreaction with
Sulphur dioxide SO2 10–105 1–9 days OH,liquid-phaseoxidation
Hydrogen sulphide H2S 0–300 ∼3 days OHMethyl mercaptan CH3SH 0.2 days OHDimethyl sulphide(DMS)
CH3SCH3 0.01–1 0.3–3 days OH, NO3,BrO
Dimethyl disulphide CH3SSCH3 ∼1day Like DMSDimethyl sulphoxide(DMSO)
CH3SOCH3 0–0.02 Days OH,deposition
Methansulphonicacid (MSA)
CH3SO3H 0–0.06 OH,gas–particleconv.
Carbon Disulphide CS2 25 ± 10 12 days OHCarbonyl sulphide COS 510 ± 70 ≈6 years OH, OSulphurous acid H2SO3 Liquid phaseSulphuric acid H2SO4 10−4–10−3 0.01 day Gas to particle
conversionSulphates, e.g.Ammoniumsulphate
(NH4)SO4
Aerosol2–8 days Deposition
Sulphur hexafluoride SF6 0.002 >1000 years Ion reactionsShort livedintermediates
Sulphur trioxide SO3 10−5 10−6 s H2OSH-radicals SH Seconds O3, O2
SO-radicals SO <3 ms O2
Deposition
OH
OHOH
OH
Aerosol
Dry depos.
CS2
SO4=
H2O
H2O2
OHSO2 SO3
H2S
COS
OH, NO3
BrO
DMS
DMSO MSA
H2SO4
Fig. 2.16. Overview of the cycles of sulphur species in the atmosphere. The surfaceocean (lower left corner) is relevant as source of atmospheric sulphur compounds
52 2 Atmospheric Chemistry
SO2 + OH M−→ HSO3 (R2.42)HSO3 + O2 → SO3 + HO2 (R2.43)
SO3 + H2OM−→ H2SO4 (R2.44)
In this reaction sequence, discovered by Stockwell and Calvert (1983), thenumber of HOX radicals is conserved, as in many other OH-initiated oxidationprocesses. Overall, OH oxidation of SO2 facilitates about one third of the totalSO2 degradation.
The oxidation of SO2 in the liquid phase also plays an important role (seeSect. 2.4.1). Gaseous SO2 is in an equilibrium with aqueous SO−
3 :
SO2(g) + H2O(liq) ↔ SO2 · H2O(liq) (R2.45)
SO2 · H2O(liq) ↔ HSO−3 (liq) + H+
(liq) (R2.46)
HSO−3 ↔ SO−
3 + H+ (R2.47)
HSO3− and SO3
− can be further oxidised to H2SO4 or sulphate, for exampleby hydrogen peroxide. This oxidation mechanism can be summarised as:
HSO3− → . . . → SO4
= (R2.48)
The sulphate formed is eventually transported to the ground together withthe (cloud or fog) droplets, thus contributing to ‘acid rain’ or ‘acid haze’.
Finally, SO2 can also react directly with the ground. This process is termeddry deposition (in contrast to processes where precipitation events play a role).About 30–50% of the SO2 is removed from the atmosphere by dry deposition.
2.7.2 Reduced Sulphur Species: DMS, COS, CS2, H2S
Most sulphur compounds, with the exception of SO2, are emitted in a re-duced form and are rapidly oxidised in the atmosphere. With the exceptionof volcanic SO2 all natural sulphur emissions are in the form of the followingspecies:
Hydrogen Sulphide (H2S) is present at mixing ratios of up to 300 ppb inthe atmosphere. It is a by-product of the degradation of proteins and smellslike rotten eggs. The oxidation of H2S is initiated by reaction with OH:
H2S + OH → H2O + SH (R2.49)
The reaction pathways of the SH radical are currently unclear. While ananalogy to the OH radical should exist, (cf. sulphur ↔ oxygen), analogousreactions of SH with methane and CO are endothermic (Becker et al., 1975b).The reaction of SH with molecular oxygen is assumed to be very slow. How-ever, SH reacts with ozone, forming SO radicals. The SO radicals finally reactwith O2 or ozone to form SO2.
2.7 Tropospheric Chemistry of Sulphur Species 53
Carbonyl Sulphide is, to a large extent, produced by the photolysis ofsulphur-containing organic species (e.g. of amino acids) in the surface layerof the ocean (e.g. Ulshofer and Andreae, 1998). Once released to the water,the majority of COS is degraded by hydrolysis to H2S (timescale of about 10–12 h). The remaining COS is emitted to the atmosphere. In addition, thereis some emission from soils and marshes (Ulshofer, 1995). In addition to thedirect emission, degradation of DMS and CS2 by OH radicals is a source ofCOS (Chin and Davis, 1993):
CS2 + OH → SCSOH (R2.50)SCSOH → COS + SO2 + H (R2.51)
Carbonyl sulphide is removed from the atmosphere by plant uptake and reac-tion with OH. The lifetime of COS in the troposphere is on the order of 2–3years, and thus only small deviations from its average mixing ratio of 500 pptare found (Ulshofer, 1995).
Vertical transport of COS plays an important role as a sulphur sourcein the stratosphere (much like N2O for the transport of oxidised nitrogen,CFCs for chlorine, and halons for Bromine, see Sect. 2.10.2). The degradationof COS in the stratosphere partially sustains the stratospheric sulphate –aerosol layer [named after its discoverer Christian Junge, ‘Junge-layer’ (Jungeet al., 1961)].
Carbon Disulphide (CS2) is naturally released via volcanoes, the ocean,marshes and forests, and industrial activities (Watts, 2000). Atmospheric mix-ing ratios are on the order of 15–35 ppt. Recently, CS2 was found by DOAS inthe city of Shanghai at levels up to 1.2 ppb (Yu et al., 2004). CS2 is degradedin the atmosphere by OH oxidation (R2.50).
Dimethyl Sulphide (CH3SCH3) is produced by biological processes in theocean (Andreae et al., 1985). Besides sporadic volcanic eruptions, oceanicDMS emissions are the largest natural source of sulphur in the atmosphere.Due to the important role of sulphur in the formation of aerosol particles andcloud condensation nuclei (CCN) DMS has received considerable attention(see also Sect. 2.7.3).
The degradation mechanisms of this species are not fully elucidated todate. Reactions of free radicals are probably largely responsible for its degra-dation. A role of OH in the degradation is likely. The first step in OH initiateddegradation of DMS is OH abstraction from one of the methyl groups, or OHaddition to the sulphur atom (Yin et al., 1990a,b).
OH + CH3SCH3 → H2O + CH3SCH2 (R2.52)OH + CH3SCH3 + M → CH3S(OH)CH3 + M (R2.53)
Intermediate products in this reaction chain are dimethyl sulphoxide(DMSO), CH3SOCH3, and CH3SO2CH3. Stable end products are sulphuricacid and methane sulphonic acid (CH3SO3H). The involvement of NO3 rad-icals has also been suggested (Winer et al., 1984; Platt and Le Bras, 1997),
54 2 Atmospheric Chemistry
where the H abstraction channel appears to be predominant (Butkovskayaand Le Bras, 1994):
NO3 + CH3SCH3 → HNO3 + CH3SCH2 (R2.54)
In addition, there are several reports of a possible role of halogen oxide radi-cals, in particular of BrO (Toumi, 1994). The product of the BrO–DMS reac-tion is DMSO:
BrO + CH3SCH3 → Br + CH3SOCH2 (R2.55)
While sulphuric acid and methane sulphuric acid form particles, DMSO doesnot. Thus, the fraction of DMS degraded by BrO may determine the efficiencyof particle formation in marine areas, as discussed by von Glasow et al. (2004).
2.7.3 Influence of Sulphur Species on the Climate,the CLAW Hypothesis
The low vapour pressure sulphur compounds, H2SO4 and methane sulphonicacid, will quickly condense after their formation in the gas phase (e.g. byreaction R2.44). In this process, small particles are formed, which can act,under appropriate conditions, as cloud condensation nuclei (CCN) and thusimpact cloud formation. Therefore, biological emission of sulphur species couldaffect cloud formation and thus our climate. This theory was proposed in1987 by Charlson, Lovelock, Andreae and Warren (referred to as the CLAWhypothesis after the authors, initials) (Charlson et al., 1987; Bates et al.,1987). The authors postulate the following mechanism: Due to microbiologicalactivities in the ocean, volatile sulphur species (essentially DMS) are released.DMS is oxidised by photochemical processes in the atmosphere to sulphuricacid or methane sulphonic acid. Due to their low vapour pressure, these speciescondense to form small particles (mean radius ∼0.07 μm), which can act asCCN.
In large areas of the worlds oceans, the supply of CCN is so small thatcloud formation can be limited by CCN availability and not by the watervapour supply (about 200CCNcm−3 are required, corresponding to a DMSmixing ratio of less than 100 ppt). The extent of cloud cover now feeds backon the insulation and thus the ocean surface temperature. If the biologicalDMS formation is positively correlated with the ocean surface temperature,which is likely, this process could form a biological negative feedback loopwhich effectively stabilises the surface temperature of earth [see also GaiaHypothesis (Lovelock, 1979)]. Since its publication, the CLAW hypothesishas been under intense debate. For instance, Schwartz (1988) argued that theemission of gaseous sulphur species in the northern hemisphere (but not inthe southern hemisphere) largely increased during the last 100 years due toanthropogenic activities (in fact the present anthropogenic S emission exceedsthe DMS source in the northern hemisphere), without major change in cloudalbedo or temperature (Schwartz, 1988).
2.8 Chemistry of Halogen Radicals in the Troposphere 55
2.8 Chemistry of Halogen Radicals in the Troposphere
During the last decade, significant amounts of the halogen oxides BrO, IO,OIO, and ClO were detected in the tropospheric BL by DOAS (Table 2.9).Direct and indirect evidence for Cl and Br atoms, as well as for Br2 and BrCl,was also found under certain conditions. In addition, there is growing evidencefor a BrO ‘background’ in the free troposphere. Observations were made at avariety of sites (see also Table 2.9 and Fig. 2.17):
The presence of these RHS in the troposphere has many consequences.Elevated RHS levels are associated with ozone destruction, which can lead tocomplete loss of BL ozone in the Arctic. This ‘Polar Tropospheric Ozone Hole’was the first hint for tropospheric halogen chemistry (Oltmans and Komhyr,1986; Bottenheim et al., 1986, 1990; Barrie et al., 1988; Platt and Lehrer,1997; Barrie and Platt, 1997; Platt and Honninger, 2003). In addition, otherdisturbances of tropospheric chemistry can occur, as detailed in Sect. 2.8.3.
2.8.1 Tropospheric Sources of Inorganic Halogen Species
The sources of RHS (X, X2, XY, XO, HOX, XONO2, HX, where X = Cl, Br,I) in the troposphere are the degradation of organic halogen compounds andthe volatilisation of halogen ions (X−) from sea salt aerosol or surface saltdeposits.
Fully halogenated compounds (such as CF2Cl2 or CF2ClBr), which are themain source of RHS in the stratosphere, are photolytically stable in the tropo-sphere. RHS can therefore only be released from less stable precursors (Cauer,1939), such as partially halogenated organic compounds like CH3Br, or poly-halogenated species, such as CHBr3, CH2Br2, CH2I2, CH2BrI (Cicerone,1981; Schall and Heumann, 1993; Khalil et al., 1993; Schauffler et al., 1998;Carpenter et al., 1999, 2001), or even I2 (Saiz-Lopez et al., 2004b; Peters,2004; Peters et al., 2005). While some of these species, such as CH3Br, areemitted by up to 50% from anthropogenic sources, most of them, and inparticular polyhalogenated species, are only emitted from biological sourcespredominately in the ocean or in coastal areas.
The atmospheric lifetime of halogenated organics varies widely. The life-time of methyl bromide, CH3Br, is on the order of 1 year, bromoform, CHBr3,has a lifetime of several days, while diiodo methane CH2I2 is photodegradedin minutes. (Wayne et al., 1995; Yvon and Butler, 1996; Davis et al., 1996;Carpenter et al., 1999). It is possible that even less stable halocarbon speciesare emitted, which, probably due to their instability, have escaped detection(Carpenter et al., 1999).
The release of inorganic halogen species from sea salt (e.g. sea salt aerosolor sea salt deposits), appears to proceed by three main pathways:
(1) Strong acids can release HCl (but not HBr) from sea-salt halides. Undercertain conditions (see below), HX can be heterogeneously converted toRHS.
56 2 Atmospheric Chemistry
Table
2.9
.O
bse
rvati
on
ofre
act
ive
halo
gen
spec
ies
inth
etr
oposp
her
eand
thei
rpro
bable
sourc
em
echanis
m
Spec
ies
Mea
sure
men
tsi
teTec
hniq
ue
Conc.
level
Ref
eren
ce
Br
Arc
tic
BL
Hydro
carb
on
Clo
ck(1
–10)·
107cm
−3
Jobso
net
al.
(1994),
Ram
ach
eret
al.
(1997,1999)
BrO
Arc
tic
and
Anta
rcti
cB
LD
OA
S(g
round
base
d)
Up
to30
ppt
Hausm
ann
and
Pla
tt(1
994),
Tuck
erm
ann
etal.
(1997),
Heg
els
etal.
(1998),
Mart
ınez
etal.
(1999),
Honnin
ger
and
Pla
tt(2
002),
Fri
eßet
al.
(2004a),
Honnin
ger
etal.
(2004c)
BrO
Arc
tic
and
Anta
rcti
cB
LD
OA
S(s
ate
llit
e)A
round
30
ppt
(ass
um
ing
1000
mla
yer
)
Wagner
and
Pla
tt(1
998),
Ric
hte
ret
al.
(1998),
Heg
els
etal.
(1998),
Mart
inez
etal.
(1999),
Wagner
etal.
(2001b),
Hollw
edel
etal.
(2004)
BrO
Salt
lakes
(Dea
dSea
,Salt
Lake
City,
Casp
ian
Sea
,Sala
rde
Uyuni)
Act
ive
DO
AS,
MA
X-D
OA
SU
pto
176
ppt
Heb
estr
eit
etal.
(1999),
Matv
eev
etal.
(2001),
Stu
tzet
al.
(2002),
Wagner
etal.
(2001b),
Honnin
ger
etal.
(2004b)
BrO
Mid
-Lat.
Mari
ne
BL
MA
X-D
OA
SU
pto
2ppt
Les
eret
al.
(2002),
Saiz
-Lopez
(2004a)
BrO
Mid
-Lat.
Fre
eTro
posp
her
eD
OA
S(d
iffer
ence
)1–2
ppt
Fri
eßet
al.
(1999),
van
Rooze
ndael
etal.
(2002),
Pundt
etal.
(2002)
BrO
Pola
rfr
eeTro
posp
her
eA
irborn
eD
OA
S0–20
ppt
McE
lroy
etal.
(1999)
2.8 Chemistry of Halogen Radicals in the Troposphere 57
.
BrO
Volc
anic
plu
mes
MA
X-D
OA
SU
pto
1000
ppt
Bobro
wsk
iet
al.
(2003),
Bobro
wsk
i(2
005)
HO
Cl/
Cl 2
Mari
ne
BL
Mis
tC
ham
ber
up
to254
ppt
Psz
enny
etal.
(1993)
Cl
Arc
tic
BL
Hydro
carb
on
Clo
ck(1
–10)·
104cm
−3
Jobso
net
al.
(1994),
Ram
ach
eret
al.
(1997,1999)
Cl
Arc
tic
BL
12C
/13C
inC
O(1
–10)·
104cm
−3
Rock
mann
etal.
(1999)
Cl
Rem
ote
Mari
ne
BL
Hydro
carb
on
Clo
ck(1
–15)·
103cm
−3
Win
gen
ter
etal.
(1996)
Cl
Extr
a-t
ropic
al
S.hem
ispher
e
12C
/13C
inC
H4
2.6−
18×
103
Pla
ttet
al.
(2004)
ClO
Salt
Lake
City
DO
AS
Up
to15
ppt
Stu
tzet
al.
(2002)
ClO
Volc
anic
plu
mes
MA
X-D
OA
S
Br 2
,B
rCl
Arc
tic
BL
AP
CIM
SU
pto
30
ppt
Fost
eret
al.
(2001),
Spic
eret
al.
(1998,2002)
I 2M
id-L
at.
Mari
ne
BL
Act
ive
DO
AS
Up
to90
ppt
Saiz
-Lopez
(2004b),
Pet
ers
(2004),
Pet
ers
etal.
(2005)
Cl 2
/H
OC
lB
r 2/H
OB
rA
rcti
cB
LC
l X≤
100
Br X
≤38
ppt
Impey
etal.
(1999)
IO,O
IOC
oast
alA
reas
DO
AS
Up
to6
ppt
(≈0.2
ppb
h−
1)
Alick
eet
al.
(1999),
Allan
etal.
(2005a),
Wit
trock
etal.
(2000a),
Fri
eßet
al.
(2001),
Saiz
-Lopez
etal.
(2005),
Zin
gle
rand
Pla
tt(2
005)
58 2 Atmospheric Chemistry
BrO, ClO BrO, IO
Sea ice Land Coast
Salt lake Open ocean
Org. precursors XO HX
Heterogeneous reactions (Aerosol)
Stratospheric BrO, ClO
Tropopause
Top of boundary layer
Fig. 2.17. The occurrence of reactive halogen species (XO, X = Cl, Br, I) in theatmosphere: BrO is found above polar sea ice during springtime, IO (and BrO) arepresent at most coastal areas. It is suspected that there is a layer of BrO in the freetroposphere (see von Glasow et al., 2004). Drawn arrows: emission of RHS, dashedarrows: emission of organohalogens
(2) Oxidising agents may convert Br− or Cl− to gaseous Br2 or BrCl. Inparticular, HOX (i.e. HOBr and HOCl) is such an oxidant. Direct photo-chemical (Oum et al., 1998a) or non-photochemical oxidation of halidesby O3 may also occur (Oum et al., 1998b; Hirokawa et al., 1998), thoughprobably quite slowly.
(3) Oxidised nitrogen species, in particular N2O5 and NO3, and perhaps evenNO2, can react with sea-salt bromide or chloride to release HBr, HCl,or photolabile species, in particular ClNO2 and BrNO2 (Finlayson-Pittsand Johnson, 1988; Finlayson-Pitts et al., 1990; Behnke et al., 1993, 1997;Rudich et al., 1996; Schweitzer et al., 1999; Gershenzon et al., 1999).
Because these mechanisms are closely linked to the atmospheric cycling ofRHS, we will discuss their details in the next section.
2.8.2 Tropospheric Cycles of Inorganic Halogen Species
In this section, we review the reaction cycles of inorganic halogen species inthe troposphere (see Fig. 2.18), (see Wayne et al., 1995; Platt and Janssen,1996; Platt and Lehrer, 1997; Lary et al., 1996; Platt and Honninger, 2003;von Glasow Crutzen, 2003).
Following release, inorganic and organic halogen species are photolysed toform halogen atoms which then predominately react with ozone:
X + O3 → XO + O2 (R2.56)
2.8 Chemistry of Halogen Radicals in the Troposphere 59
hν
OXO
X2,XNO2
XY
X2O2XONO2
HX
CH3X,CHX3,etc.
Y –
N2O5 , etc.
heterogeneous
hν, Δ
hνhν
hν
hν,Δ
OH, hν
RH,HO2
HO YO O3HO2
NO2
YO
hνHOX
XO
X
XO
(X2O2)ads(XONO2)ads (HOX)ads(HX)ads
Sea Salt (Snow pack or aerosol surface)
H+ + X–H2O
Fig. 2.18. Reaction cycles of reactive halogen species in the troposphere
Typical lifetimes of halogen atoms (X) due to R2.56 at tropospheric back-ground O3 levels are around 0.1 s for Cl, and on the order of 1 s, for Br and I.Halogen atoms are regenerated in a series of reactions including photolysis ofXO, which is of importance for X = I, Br, and, to a minor extent, Cl:
XO + hν → X + O (R2.57)
where J2.57 ≈ 3× 10−5 s−1, 4× 10−2s−1, 0.2s−1 for X = Cl, Br, and I, respec-tively. The reaction of XO with NO also recycles halogen atoms:
XO + NO → X + NO2 (R2.58)
This reaction leads to a shift of the NO/NO2 photostationary state inthe atmosphere, by providing a shortcut in the normal NO/NO2/O3 cycle(R2.1/R2.2/R2.3). This has consequences for the photochemistry in the tro-posphere and, in particular, ozone formation (Stutz et al., 1999).
The self-reactions of XO (or reaction with another halogen oxide YO) alsoplay a role:
XO + YO → X + Y + O2 (R2.59a)XO + YO → XY + O2 (R2.59b)XO + YO → OXO + O (R2.59c)
XO + YO M−→ X2O2 (R2.59d)
60 2 Atmospheric Chemistry
Reactions R2.56, R2.59a,b,d, and the photolysis of the halogen moleculesformed in R2.59 constitute a catalytic cycle which destroys ozone in the tro-posphere. The efficiency of this cycle is determined by the amount of RHSpresent and the rates of R2.59a,b,d. This type of halogen-catalysed ozone de-struction has been identified as the prime cause of polar BL ozone destruction(Oltmanns and Komhyr, 1986; Barrie et al., 1988; Barrie and Platt, 1997).
Halogen atoms can also react with saturated (Cl) or unsaturated HC (Cland Br) to form hydrogen halides, e.g.:
RH + Cl → R + HCl (R2.60)
Here, R denotes an organic radical. An alternative is the reaction of halogenatoms with formaldehyde (or higher aldehydes) or HO2, also leading to theconversion of halogen atoms to hydrogen halides:
X + HO2 → HX + O2 (R2.61)X + HCHO → HX + CHO (R2.62)
The former reaction occurs for X = Cl, Br, I, the latter only for X = Cl, Br. Inparticular, Cl is a very strong oxidant and a Cl concentration of 104 cm−3 canalready considerably contribute to the oxidation capacity of the troposphere.
The main loss of RHS is their conversion to hydrogen halides which arehighly water soluble, and can thus be easily lost from the atmosphere by wetor dry deposition. The only relevant gas-phase ‘reactivation’ mechanism ofHX is reaction with OH:
HX + OH → X + H2O (R2.63)
Only at high HO concentration and in the absence of deposition processes,such as in the free troposphere, does this reaction play a role. HX can also berecycled through the aerosol phase, as we will discuss below.
Other important inorganic halogen species are HOX and XONO2. HOX isformed via the reaction of peroxy radicals with XO (Canosa-Mas et al., 1999):
XO + HO2 → HOX + O2 (R2.64)
Formation of HOX is followed by its photolysis:
HOX + hν → X + OH (R2.65)
Reactions R2.56, R2.64, and R2.65 form another catalytic cycle that destroysozone during the day. This cycle dominates at low RHS levels, such as thosefound in the marine BL. The catalytic ozone destruction cycle through R2.59 isimportant at higher RHS levels, such as those found in polar regions. ReactionsR2.64 and R2.65 also lead to a shift in the HO2/OH ratio in the troposphere,
2.8 Chemistry of Halogen Radicals in the Troposphere 61
and thus directly influence the oxidising capacity of the troposphere (Blosset al., 2005).
Another important reaction cycle involving HOBr is the liberation ofgaseous bromine species (and to a lesser extent chlorine species) from (sea-salt) halides (Fan and Jacob, 1992; Tang and McConnel, 1996; Vogt et al.,1996):
HOBr + (Br−)Surface + H+ → .Br2 + H2O (R2.66)
The required H+ [the reaction appears to occur at appreciable rates only atpH < 6.5 (Fickert et al., 1999)] can be supplied by strong acids, such as H2SO4
and HNO3 from anthropogenic or natural sources. Reaction R2.66, togetherwith R2.64, R2.56, and the photolysis of Br2, form a cycle where Br−, forexample, in sea ice or the aerosol, is volatilised. Because the uptake of oneHOBr molecule leads to the formation of two Br atoms, which, ignoring anyBr loss, can form two HOBr molecules (R2.64), this cycle is autocatalytic. Theexplosion-like behaviour of this cycle has lead to the term ‘Bromine Explosion’for this cycle, which is believed to be responsible for the high reactive brominelevels found in the Arctic (Platt and Lehrer, 1995; Platt and Janssen, 1996;Wennberg, 1999).
The reaction of halogen oxides with NO2 forms halogen nitrate, XONO2:
XO + NO2M−→ XONO2 (R2.67)
Bromine nitrate is assumed to be quite stable against thermal decay, but isreadily photolysed and may be converted to HOX by heterogeneous hydrolysis(van Glasow et al., 2002):
XONO2 + H2OM−→ HNO3 + HOX (R2.68)
or to Br2 or BrCl by heterogeneous reaction with HY:
XONO2 + HY Surface−→ HNO3 + XY (R2.69)
Overall, XONO2 is probably of minor importance at the low NOX levelstypically found in the free troposphere, but it can play an important rolein polluted air, for instance at polluted coastlines. There XONO2 can be animportant source of HOX, thus in particular BrONO2 might contribute tobromine explosion events.
Under conditions of high NOX , halogen release can also occur via thereactions:
N2O5(g) + NaX(s) → NaNO3(s) + XNO2(g) (R2.70)
(e.g. Finlayson-Pitts et al., 1989). The XNO2 formed in the above reactionmay photolyse to release a halogen atom, or possibly further react with seasalt (Schweitzer et al., 1999):
XNO2(g) + NaX(s) → NaNO2(s) + X2(g) (R2.71)
62 2 Atmospheric Chemistry
The above reaction sequence would constitute a dark source of halogenmolecules. The direct reaction of NO3 on salt surface can also release reactivehalogens, for example in the form of halogen atoms:
NO3(g) + NaX(s) → NaNO3(s) + X(s) (R2.72)
(Seisel et al., 1997; Gershenzon et al., 1999). The uptake coefficient of NO3
by aqueous solutions of NaBr and NaCl was found by Rudich et al. (1996) tobe near 0.01 for sea water, while Seisel et al. (1997) found up to 0.05 for dryNaCl.
2.8.3 Potential Impact of Inorganic Halogen Specieson Tropospheric Chemistry
The role of reactive halogens in the troposphere has been a controversial issueand remains a poorly understood aspect of atmospheric chemistry. However,it is becoming clear that the RHS are present in many parts of the troposphere(Table 2.9). At the levels shown in Table 2.9, RHS can have a noticeable effecton several aspects of tropospheric chemistry. These include:
(1) RHS, and in particular reactive bromine and iodine can readily destroytropospheric ozone through catalytic cycles (Solomon et al., 1994b; Plattand Janssen, 1996; Davis et al., 1996; Sander et al., 2003; von Glasow et al.,2004). This has consequences for atmospheric chemistry and, consideringthat ozone is a greenhouse gas, also the global climate (e.g. Roscoe et al.,2001).
(2) Reactions R2.64, and J2.65 will lead to the conversion of HO2 to OH, andthus reduce the HO2/OH ratio, with consequences for the atmosphericoxidation capacity and thus indirectly for the lifetime of methane, anotherimportant greenhouse gas.
(3) The presence of highly reactive chlorine atoms, and to a lesser extentbromine atoms, has a direct impact on the oxidation capacity of the tro-posphere.
(4) The reaction of BrO with DMS might be important in the unpollutedremote marine BL, where the only other sink for DMS is the reactionwith OH radicals (Toumi, 1994; von Glasow and Crutzen, 2003, 2004). Itmay thus have an impact on cloud formation as explained in Sect. 2.8.3.
(5) Deposition of mercury was found to be enhanced by the presence of re-active bromine species, in particular in polar regions (e.g. Barrie andPlatt, 1997; Schroeder et al., 1998). This process appears to be one ofthe most important pathways for toxic mercury compounds to enter thearctic (Lindberg et al., 2002), and possibly also other ecosystems.
(6) Iodine species might be involved in particle formation in the marineBL (Hoffmann et al., 2001; O’Dowd et al., 2002; Jimenez et al., 2003;Burkholder et al., 2004).
2.9 Oxidation Capacity of the Atmosphere 63
At present, more measurements are needed to ascertain the distribution andlevels of RHS in the troposphere and to assess the significance of RHS onlocal, regional, and global atmospheric chemistry and our climate.
2.9 Oxidation Capacity of the Atmosphere
The term oxidation capacity, sometimes also called oxidative power, refersto the capability of the atmosphere (or rather a part thereof) to oxidise (orotherwise degrade) trace species emitted into it. This ability is crucial for theremoval of trace species, such as the greenhouse gas methane, and is thus oftenalso referred to as the ‘self-cleaning’ capacity of the atmosphere. Althoughthere is no general definition, the oxidation capacity is frequently associatedwith the abundance of OH. However, as explained above, many other oxidants(including O2 and O3), as well as free radicals other than OH, can contributeto the oxidation capacity of the atmosphere. While the oxidation capacity isquite a popular concept, it is nevertheless difficult to define.
One difficulty in defining the term oxidation capacity lies in the fact that agiven agent (e.g. OH radicals) might act quite differently on various pollutants.Perhaps the best definition for the oxidation capacity Ci of a radical Ri (e.g.Ri = NO3, OH, O3, Cl) with the atmospheric abundance (concentration) [R],with respect to the degradation (oxidation) of species Xj (e.g. volatile organicspecies, CO, and NOX), oxidised by the molecule (per second and cm−3),might be a sum such as:
Ci =∑
j
[Xj ] · [Ri] · kij (R2.73)
where [Ri] denotes the concentration of radical species i with the reactionrate constant kij towards species Xj (Geyer et al., 2001; Platt et al., 2002).Frequently 24-h average values of the oxidation capacity are calculated.
Table 2.10 summarises typical and maximum tropospheric concentrationsof the radicals discussed, and shows minimum and typical fractional lifetimesof key VOCs against degradation by the various free radicals.
As an example, the relative contribution of the atmospheric oxidants OH,NO3, and ozone to the 24 h average of the degradation of organic species inthe atmosphere, as observed during the BERLIOZ 1998 campaign at Papst-thum near Berlin, Germany, is shown in Fig. 2.19. When all VOCs (includingCH4 and CO) are considered, OH reactions constitute about three quartersof the oxidation capacity. For NMHC, NO3 contributes almost one third ofthe oxidation capacity (Geyer et al., 2001).
Open questions include the role of heterogeneous reactions in radical cy-cles. While the contribution of heterogeneous reactions to HOX chemistry isunclear to date, it is likely that heterogeneous reactions play an important rolein the atmospheric cycles of most other radicals. Another interesting question
64 2 Atmospheric Chemistry
Table
2.1
0.
Est
imate
dlife
tim
e(h
ours
)fo
rse
lect
edV
OC
Spec
ies
OH
(av.)
OH
(pea
k)
NO
3(a
v.)
NO
3(p
eak)
ClO
(av.)
ClO
(pea
k)
Cl(a
v.)
Cl(p
eak)
BrO
(av.)
Br
(av)
Conc.
mole
c.cm
−3
6×
105
1×
107
1.5
×108
1×
1010
106
108
103
105
108
5×
106
DM
S110
6.3
1.7
0.0
33×
104
280
870
8.7
10c
26
n-B
uta
ne
190
11
2.8
×104
430
–1400
14
C2H
456
3.4
1×
104
160
2800
28
Isopre
ne
4.6
0.3
1.9
0.0
3–
<1100
<11
0.9
Tolu
ene
77
4.6
3.1
×104
460
–C
H4
7×
104
4×
103
>5×
104
>700
>7×
107
>7×
105
3×
106
3×
104
2.10 Stratospheric Ozone Layer 65
is whether the effect of several oxidising species is additive. In fact, it is possi-ble that the presence of halogen oxides may reduce the OH concentration dueto an increase in the photo-stationary state NO2/NO ratio. In addition, it iswell possible that there are other, yet unrecognised, mechanisms involving theabove or other free radicals.
2.10 Stratospheric Ozone Layer
The role of atmospheric ozone as a filter for the Sun’s harmful UV radiationhas lead to an early interest in ozone chemistry in the investigation of thechemistry of the atmosphere. Spectroscopic measurements by Gordon MillerBourne Dobson in 1925 showed an ozone column density larger than expectedfrom assuming the O3 mixing ratio found near the surface to be constantthroughout the atmosphere (e.g. Dobson and Harrison, 1926; Dobson, 1968).Subsequent spectroscopic measurements using the so-called ‘Umkehr’ effectby Paul Gotz proved the theory of an ozone layer and determined the alti-tude of its maximum at 25 km (Gotz et al., 1934). This stratospheric ozonelayer contains 90% of the atmosphere’s ozone and is thus responsible for themajority of the absorption of solar UV radiation.
2.10.1 Stratospheric Ozone Formation: The Chapman Cycle
In the late 1920s, Sidney Chapman proposed a reaction scheme which ex-plained the observed vertical profile of ozone, with relatively low mixing ra-tios in the troposphere and a maximum around 25 km altitude, the ‘ChapmanMechanism’ (Chapman, 1930). The initial process is the photolysis of oxy-gen molecules to form two oxygen atoms in their ground state (indicated bythe spectroscopic notation (3P), see Chap. 3). In the stratosphere, sufficientlyenergetic UV light (i.e. light with wavelengths below 242 nm) is available tophotolyse oxygen molecules:
O2 + hν(λ < 242 nm) → O(3P) + O(3P) (R2.74)
The oxygen atoms can react in three ways: (1) Recombine with an oxygenmolecule to form ozone. Since two particles (O and O2) combine to make one(O3), collision with a third body (M, likely N2 or O2,) is required to facilitatesimultaneous conservation of energy and momentum. The reaction is thereforepressure dependent:
O(3P) + O2 + M → O3 + M (R2.2)
Alternatively, (2) the oxygen atom can react with an existing ozone molecule:
O(3P) + O3 → O2 + O2 (R2.74)
66 2 Atmospheric Chemistry
Fig. 2.19. The relative contribution of the atmospheric oxidants OH, NO3, andozone to the degradation of organic species in the atmosphere on a 24-h basis. Casestudy during the BERLIOZ 1998 campaign at Papstthum near Berlin, Germany.(a) VOCs including CH4 and CO, (b) non-methane VOCs, (c) alkenes only (fromGeyer et al., 2001, Copyright by American Geophysical Union (AGU), reproducedby permission of AGU)
Finally, (3) the recombination of two oxygen atoms to form molecularoxygen is possible but largely unimportant in the stratosphere:
O(3P) + O(3P) + M → O2 + M (R2.75)
2.10 Stratospheric Ozone Layer 67
In addition to the ‘primary’ production of O atoms by the photolysis of O2,the photolysis of O3 also provides ‘secondary’ O atoms. In fact, photolysis ofozone molecules occurs at a much higher rate than that of oxygen molecules:
O3 + hν(λ < 612 nm) → O(3P) + O2 (R2.76)
In summary, the above reactions, also known as the ‘Chapman Reactions’,lead to a steady-state O3 level in the stratosphere, in which the O atom pro-duction via reactions R2.74 and R2.76 is in balance with their destruction viarecombination with O2 and reaction with O3. The above set of reactions ex-plains the formation of a layer of ozone with a maximum concentration in thelower stratosphere. In the lower stratosphere and troposphere, the rate of O2
photolysis, and thus the ozone formation rate, becomes extremely low (despitethe much higher O2 concentration there). However, O3 destruction still occursvia O3 photolysis, which takes place at much longer wavelengths, and the re-action of O + O3. This explains why the ozone concentration should increasewith height (in fact the Chapman mechanism predicts zero O3 formation inthe troposphere). On the other hand, in the upper part of the stratosphere therecombination of O + O2 (reaction R2.2) becomes slower, since the concen-tration of air molecules necessary as a ‘third body’ (M) in the recombinationof O + O2 (reaction R2.2) reduces proportionally to the atmospheric pres-sure. Thus, despite increasing levels of UV radiation, the O3 concentration(and also the mixing ratio) will eventually decrease with altitude. Figure 2.20depicts the ozone profile predicted by the Chapman cycle.
2.10.2 Stratospheric Ozone Chemistry: Extensionof the Chapman Cycle
The Chapman mechanism gave a satisfactory explanation for the occurrenceof an ozone layer. However, detailed investigation during the 1960s of theelementary reactions and photolysis processes involved revealed that quanti-tatively the mechanism overestimates the O3 levels by about a factor of three,see Figure 2.20. It subsequently became clear that there are many other tracegas cycles affecting stratospheric O3 levels. In particular, a group of reactionswere found to catalyse the elementary reaction of O + O3 (reaction R2.74above). These reaction sequences follow the general scheme:
O3 + Z → ZO + O2 (R2.77)
ZO + O(3P) → Z + O2 (R2.78)
With the net result,O(3P) + O3
Z−→ O2 + O2 (R2.79)
where Z (and ZO) denotes a species acting as catalyst Z for reaction R2.75. Inthe stratosphere, the following species are active (see Table 2.11): Cl (ClO),
68 2 Atmospheric Chemistry
010
20
30
40
5 10 15 20
Ozone mixing ratio (ppm)
Measurements
Chapman cycle
Hei
ght (
km)
Fig. 2.20. The stratospheric ozone profile according to the Chapman Mechanism,compared to observations (figure adapted from Roth, 1994)
Br (BrO), NO (NO2), or OH (HO2). Inclusion of these reactions brings obser-vations and model calculations in very good agreement. Figure 2.21 illustratesthe contribution to the ozone destruction of each of the catalysts at differentaltitudes.
All species listed in Table 2.11 play an important role in stratosphericchemistry. The oxides of nitrogen (NO, NO2, see Sect. 2.5) are formed in thestratosphere by reaction of excited oxygen atoms O(1D) with N2O (whichcan be regarded as a ‘transport species’ for oxidised nitrogen into the strato-sphere):
N2O + O(1D) → NO + NO (R2.80)→ N2 + O2
Table 2.11. ‘Catalysts’ destroying stratospheric ozone
Catalyst Z ZO Stratospheric source
NO NO2 Degradation of N2O from thetroposphere
OH HO2 O1D + H2OCl ClO Degradation of (mostly man-
made) organo chloridesBr BrO Degradation of CH3Br, and other
organo bromine species
2.10 Stratospheric Ozone Layer 69
0
10
20
30
40
20 40
OxNOx
O3 Loss rates
HOx
CIOx + BrOx
Z (
km)
60
% of Total Ox loss
80 100
Fig. 2.21. The contribution of the ClOX , BrOX , NOX , and HOX , catalysed de-struction of stratospheric ozone to the original Chapman Mechanism as a functionof altitude (from Portmann et al., 1999, Copyright by American Geophysical Union(AGU), reproduced by permission of AGU)
The excited oxygen atoms O(1D) are formed in the UV photolysis of O3 by(R2.4). As in the troposphere, a photo-stationary state between NO, NO2,and O3 is established and conversion of NO2 to N2O5 will take place. Un-der ‘normal’ stratospheric conditions, loss of N2O5 occurs due to photolysis,leading to an increase of NOX during the course of a day (only during polarwinter do heterogeneous reactions of N2O5, e.g. R2.98 become important).
Halogen atoms are released by the degradation of long-lived halocarbonspecies (e.g. CH3Cl,CFCl3, CH3Br, etc.), which are emitted naturally or an-thropogenically in the troposphere and then transported into the stratosphere.Halogen atoms (denoted as X = F, Cl, Br, I), which are released in the degra-dation processes, react quickly with O3 to form halogen oxides (see Sect. 2.8).In addition, halogen atoms can react with O2 by association reactions to formhalogen peroxide radicals. These radicals are unstable and dissociate by thereverse of their formation reaction, so that under atmospheric conditions theXO2 species exists in equilibrium with X and O2.
X + O2(+M) → XO2(+M) (R2.81)
The stability of XO2 decreases in the series F > Cl > Br > I. The equilibriumfor atmospheric conditions lies in favour of the atom, and XO2 reactions donot play a significant role in stratospheric chemistry.
70 2 Atmospheric Chemistry
A rapid reaction occurs between XO and NO:
XO + NO → X + NO2 (R2.82)
NO2 is rapidly photolysed to produce atomic oxygen, which reforms ozone,and thus the overall effect of this reaction coupled with the X + O3 reactionon ozone is neutral. On the other hand, this reaction will lead to a higherstationary state X/XO ratio, and thus (at least in the case of chlorine) enhancethe rate of X → HX conversion by reaction of halogen atoms with HC (i.e.CH4). Consequently, the presence of nitrogen oxides tends to increase theHX/XOX ratio, and thus reduce the ozone depletion by halogen radicals.
The chlorine oxide radical undergoes self-reaction to form a dimer, Cl2O2
(Cox and Derwent, 1979). The ClO dimer plays a major part in the catalyticdestruction of ozone occurring in the polar lower stratosphere during win-tertime, which in Antarctica leads to the formation of the ozone hole (seeSect. 2.10.3 below). The relevant cycle is:
ClO + ClO + M → Cl2O2 + M (R2.83)Cl2O2 + hν → Cl + ClO2 (R2.84)ClO2 + M → Cl + O2 (R2.85)Cl + O3 → ClO + O2 (R2.86)
The ClO dimer is relatively unstable, and the cycle is only important at lowtemperatures (T ≈ 200K) when the reverse decomposition of Cl2O2:
Cl2O2 + M → ClO + ClO + M (R2.87)
is slow compared with its photolysis.Symmetric chlorine dioxide is formed in the stratosphere from the coupled
reaction of BrO with ClO:
BrO + ClO → OClO + Br (R2.88a)
The discovery of atmospheric OClO using ground-based DOAS spectroscopyat the McMurdo Sound base in Antarctica (Solomon et al., 1987b), was thefirst positive indication that partitioning of atmospheric chlorine between itsactive and inactive forms was greatly perturbed in the Antarctic ozone hole.Chlorine dioxide is rapidly photolysed by visible light to yield O atoms
OClO + hν → O + ClO (R2.89)
and thus is not a catalytic agent for O3 destruction. A consequence of its rapidphotolysis are low OClO concentrations during daytime.
Observations of the evolution of stratospheric OClO during sunset allowdeduction of information on the BrO concentration in that BrO is an essential‘catalyst’ for its formation. On the other hand, BrO is lost by BrCl productionvia reaction:
2.10 Stratospheric Ozone Layer 71
BrO + ClO → BrCl + O2 (R2.88b)
while the third channel of reaction R2.88 does not change the XO reservoir,but leads to the destruction of ozone:
BrO + ClO → Cl + Br + O2 (R2.88c)
Since k2.88b/(k2.88a+k2.88b+k2.88c) ≈ 0.08, the late night OClO concentrationcannot become larger than 1/0.08 ≈ 12 times the daytime BrO concentration(e.g. Wahner and Schiller, 1992).
The behaviour of halogen oxides in the stratosphere is strongly influ-enced by the chemistry of temporary reservoir species containing chlorine andbromine. Temporary reservoirs are molecules which are formed and brokendown at a time scale on the order of 1 day, contain halogens, and are not ac-tive in O3 destruction. For example, chlorine nitrate is an important reservoirfor ClOx. It is formed by the reaction of ClO with NO2 (R2.67), and activechlorine is released by photolysis; which occurs mainly by the reaction:
ClONO2 + hν → Cl + NO3 (R2.90)
orClONO2 + hν → ClO + NO2 (R2.91)
The other important temporary reservoir for chlorine is HOCl, formed byreaction of ClO with HO2:
ClO + HO2 → HOCl + O2 (R2.92)
Chlorine is released from HOCl by photolysis. The major channel of thisreaction is:
HOCl + hν → HO + Cl (R2.93)
In an analogous way ClO may react with CH3O2, producing methyl-hypochlorite CH3OCl, or other products (Helleis et al., 1993, 1994):
ClO + CH3O2 → CH3OCl + O2 (R2.94a)→ Cl + products (R2.94b)
Cycles involving photolytic production of Cl from ClONO2 and HOCl can leadto depletion of ozone in the lower stratosphere, and these may be significantunder some circumstances [see for instance Toumi and Bekki, 1993]. Similarly,reaction R2.94a, followed by the photolysis of CH3OCl or reactions of typeR2.94b, could cause ozone loss (Crutzen et al., 1992).
In the polar stratosphere and in circumstances of elevated concentrationsof volcanic aerosol, heterogeneous reactions of ClONO2 and HOCl on polarstratospheric clouds (PSCs; particles containing nitric acid trihydrate and/orwater ice) lead to gross perturbation of the distribution of ClOx betweeninactive reservoirs and active ozone depleting forms. The reactions involvedare:
72 2 Atmospheric Chemistry
ClONO2 + H2O(s) surface−−−−−→HOCl + HNO3(s) (R2.95)ClONO2 + HCl(s) → Cl2 + HNO3(s) (R2.96)HOCl + HCl(s) → H2O(s) + Cl2 (R2.97)N2O5 + HCl(s) → ClNO2 + HNO3(s) (R2.98)
Here, (s) refers to the chemical species adsorbed or absorbed in the solid orliquid phase.
As bromine is less tightly bound than chlorine, only a small fraction of thebromine released from bromocarbons is sequestered in the form of HBr andBrONO2, rendering this atom very effective for ozone loss (e.g. Lary, 1995). Inparticular, the combined Br–Cl catalytic cycles are very efficient in depletingozone, and can therefore cause equal or even larger ozone destruction thanchlorine alone in the lower stratosphere. Although there are significant humansources of bromine, the contemporary abundance of total stratospheric Br isonly about 0.5% of that of Cl (e.g. Schauffler et al., 1998; Wamsley et al.,1998). The organic precursor species of inorganic bromine, = 50%CH3Br,halons, CHBr3, etc. are still increasing. When the precursor species reachthe lower stratosphere, they are photolysed and release inorganic bromine(BrY = BrO, HBr, HOBr, BrONO2, BrONO, Br, and Br2). Due to the lowerbinding energy of HBr compared with HCl, and to the more rapid photolysisof Br2, BrCl, HOBr, BrONO2 compared with their Cl analogues, BrO is themost abundant bromine species during the daytime and constitutes between50% and 70% of total BrY. In mid-latitude summer with high stratosphericNO2 levels, BrO is less abundant (≈50% of the total BrY), while BrO canamount up to 70% of BrY at high latitudes in a denoxified stratosphere. Thisis because the BrY reservoir species BrONO2 is less abundant in a stratospherewith low NOX loading.
The possible participation of iodine chemistry in stratospheric processesis still under debate. Owing to the much shorter atmospheric lifetimes ofiodine-containing compounds (e.g. CH3I emitted from the oceans) comparedwith their Cl- and Br-containing analogues, it has generally been assumedthat iodine does not reach the stratosphere in significant quantities, and themajor focus of studies of iodine photochemistry has been the troposphere,as discussed in detail earlier. However, should iodine reach the stratosphere,its impact on ozone depletion ‘molecule for molecule’ is probably even greaterthan that of Br. The main reason for this is the inherent instability of potentialiodine reservoir species (HI, HOI, or IONO2) compared with their Cl- andBr-containing counterparts (Solomon et al., 1994). In addition, the potentialcoupling of the chemistry of IO with ClO and BrO may be far more efficientthan the XO self-reaction cycles referred to above.
2.10.3 Stratospheric Ozone Hole
In 1985, Farman et al. (1985) observed a decrease of the total ozone col-umn over Antarctica after polar sunrise. Satellite observations revealed that
2.10 Stratospheric Ozone Layer 73
this ozone depletion occurred over the entire Antarctic continent. It soonbecame clear that, during Antarctic winter, special conditions prevail whichlead to the release of reactive chlorine from its ‘reservoir compounds’ (HCland ClNO3) by heterogeneous reactions, i.e. chemical processes at the surfaceof particles in the stratosphere (see reactions R2.95–R2.98) (e.g. Solomon,1999). This massive activation, bringing the fraction of reactive chlorine fromaround 1% to near 50%, leads to dramatic ozone losses in the lower strato-sphere. A few examples are illustrated in Fig. 2.22, which shows normal ozoneprofiles recorded in August (Antarctic winter) and disturbed profiles in Oc-tober (Antarctic spring). There is nearly complete ozone loss in the altituderange from about 14 to 24 km.
A number of factors contribute to the formation of the ozone hole: Theprimary cause is the increase in stratospheric chlorine and bromine levelsdue to the release of man-made CFCs, such as CF3Cl and CF2Cl2. Thesecompounds, which have no sinks in the troposphere, have been accumulat-ing since their introduction in the 1930s. Because of their chemical stability,CFCs were widely used as propellants, in refrigerators, as solvents, and in
00
5
10
15
20
25
30
35
5
O3 Partial pressure (mPa)
Alti
tude
(km
)
10 15
21 AUG 1996
MCMURDO (78°S, 167°E)
01 OCT 199624 AUG 199725 OCT 1997
20
Fig. 2.22. Ozone profiles in August (Antarctic winter) 1996 and October (Antarcticspring) of 1997 (from Nardi et al., 1999, Copyright by American Geophysical Union(AGU), reproduced by permission of AGU)
74 2 Atmospheric Chemistry
other industrial applications. Most man-made halocarbons transported to thestratosphere contain more than one halogen atom, and the release of halogenatoms occurs in stages, with intermediate formation of the carbonyl halides(e.g. CFClO). For all halogens, the sink is provided by transport of the moststable (and therefore most abundant) forms (e.g. HX) to the troposphere.
A polar vortex forms in the winter polar stratosphere, which inhibits gas-exchange with lower latitudes. As a consequence, colder polar air cannot mixwith warmer lower latitude air, leading to very cold stratospheric tempera-tures. In addition, ozone from lower latitudes cannot replenish ozone destroyedby halogen catalysed cycles. The polar vortex is very stable over Antarctica,and can exist from June to November. The northern polar vortex is less sta-ble and is typically not stable for more than a few weeks. The more unstablenorthern vortex is the main reason why the ozone depletion over the northpole has not reached the same extent as that over the south pole.
Very cold stratospheric temperatures develop over Antarctica in winterdue to radiative cooling and lack of solar heating. In addition, the lack of airexchange with mid-latitudes also contributes to the low temperature. Oncethese temperatures fall below a certain value (∼195K), PSCs form. The cloudsare made out of nitric acid tri-hydrate and water. PSCs provide a surfaceon which the chlorine reservoir species, such as HCl and ClONO2 can beconverted to Cl2 (R2.95 and R2.96). In addition, PSC formation removesnitrogen oxides from the gas phase, further increasing the amount of reactivehalogens. These heterogeneous processes proceed for several months, as longas PSCs are present.
The rapid photolysis of Cl2 releases reactive chlorine, in the form of Clatoms, at polar sunrise. Cl reacts quickly with O3 forming ClO, which thenparticipates in the catalytic ozone destruction cycles described earlier. ClOis the most important halogen oxide in the stratosphere: its concentrationranges from 10–100 ppt in the undisturbed lower stratosphere to 1–2 ppb underconditions of disturbed stratospheric chemistry, i.e. the ozone hole, where morethan 50% of the available chlorine can be present as ClO. Under the conditionsof very high ClO levels and low temperatures, the ClO self-reaction to forma dimer, Cl2O2 [reaction R2.83 above (Cox and Derwent, 1979; Molina andMolina, 1987)] becomes the dominant catalytic destruction cycle of ozone,leading ultimately to the formation of the ozone hole. It should be noted thatcycles involving BrO also contribute to the ozone destruction, although toa lesser extend, since the partitioning between active and reservoir speciesfavours the active forms already in the undisturbed stratosphere.
The break-up of the polar vortex in December marks the disappearanceof the ozone hole, as warmer and ozone-rich air masses are transported overAntarctica.
The formation of the ozone hole is a recurring phenomenon, which startsevery Antarctic winter.
2.10 Stratospheric Ozone Layer 75
Fig. 2.23. Time series of atmospheric chlorine loading from 1960 – today, with pro-jections to 2080. The expected effect of the treaties for limitation of CFMs is shown(from Brasseur and Solomon 1986)
2.10.4 Recovery of Stratospheric Ozone
Recently, there have been signs of recovery of stratospheric ozone, which islargely due to the successful measures employed to reduce the release of halo-gen transport species (e.g. CFCs).
Figure 2.23 shows time series of the atmospheric chlorine loading, thetotal amount of chlorine present as transport, reservoir, and active speciesfrom 1960 to 1995 (measured data), and projections to 2080.
The chlorine loading of the troposphere (Montzka et al., 1996) and strato-sphere (Rinsland et al., 2003) peaked in 2004, and has begun to decline. Thefirst indication of the expected recovery of stratospheric ozone may alreadybe visible (Newchurch et al., 2003; Bodeker et al., 2005).